{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Predicting Uncertainty:\n", "\n", "\n", "**Prediction Intervals from Gradient Boosting Quantile Regression**\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "![image](https://i.imgur.com/3Tilb52.jpg)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "\n", "> What is the difference between actuaries and fortune tellers?\n", ">\n", "> Answer: Fortune tellers give you point estimates, but actuaries give you *prediction intervals*. Both try to predict the future.\n", "\n", "Models continue to become more accurate, but for certain applications, the uncertainty of an estimate is as important as the prediction itself. By using prediction intervals, which create a range of values instead of a single number, this uncertainty can be measured using quantile regression. In insurance and finance where model risk translates to financial gain or loss, predictive analytics practitioners can reduce losses and increase profits by hedging against this risk. \n", "\n", "Using gradient boosting, a widely-used machine learning method, two use cases are demonstrated on open-source and reproducible data:\n", "\n", "1. Add value to actuarial predictive models by creating prediction intervals;\n", "2. Audit data quality in a fully-scalable manner using outlier detection." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Introduction\n", "---\n", "\n", "Predictive analytics is the science of using statistics to predict the future. This processes often involves identifying a business problem, collecting relevant data, and then making predictions that help to improve or solve the problem. As advances in the field of machine learning produce better algorithms, the relative size of errors continues to decrease. But even the best models are random to some degree. \n", "\n", "In real-world applications, a prediction about a future event will never completely match reality. Even with the most robust validation, we can never know for certain how well a given model will perform on unseen data. Data collected outside of the laboratory is often sparse for certain segments of the population. In health care, models are built based on thousands of diagnosis codes, and each patient may only have a few dozen codes in common with other patients. Credit unions and banks create models from purchase histories which contain varying line item keywords. Websites use clickstream data which contains millions of user inputs. In all of these examples, the data quality can vary dramatically from case-to-case. Intuitively, those predictions which are based on low-quality data will be less reliable than those with more data.\n", "\n", "For financial applications where the prediction is a dollar amount, model risk translates into financial risk. In [Modern Portfolio Theory (MPT)](https://www.investopedia.com/terms/m/modernportfoliotheory.asp), investment risk is measured by the standard deviation. This is the tendency of the asset to have a return that is above or below the average return. Portfolio managers need to know the risk level as well as the average return in order to maintain profitability.\n", "\n", "In insurance, actuaries are the original scientists of risk. Traditionally, actuaries use parametric models which allow for not only the average to be estimated, but also other [risk measures](https://www.casact.org/library/studynotes/hardy4.pdf). The Value at Risk (VaR) is a terminological name for the pth quantile of the loss distribution. Usually this is just the 90th or 99th percentile. Two entities with the same average but different VaR are valued very differently.\n", "\n", "The Actuarial Standards of Practice (ASOPs) describe the procedures an actuary should follow when performing actuarial services. ASOP No. 23 speaks to this issue of data quality and how the uncertainty behind results should be acknowledged.\n", "\n", ">Appropriate data that are accurate and complete may not be available. The actuary should use available data that, in the actuary’s professional judgment, allow the actuary to perform the desired analysis. *However, if significant data limitations are known to the actuary, the actuary should disclose those limitations and their implications in accordance with section 4.1(b).* \n", "> - [ASOP No. 23 § 3.1 (emphasis added)](http://www.actuarialstandardsboard.org/asops/data-quality/).\n", "\n", "Much research has been done to create machine learning models which create more accurate averages, but less research focuses on measuring the uncertainty of those predictions. One of the most widely-applicable algorithms, the gradient boosting machine (GBM), has a method of estimating prediction error using quantile regression. This allows for predictions of not only the average, but also the range or uncertainty of the prediction. This helps to solve the problem of model risk: if an observation is predicted based on sparse data, then the prediction interval for this record will be wide; if the prediction is based on strong evidence, or in actuarial terminology has \"higher credibility\", then the prediction interval will be narrower.\n", "\n", "### Quantile Regression\n", "\n", "Most models estimate averages, but quantile regression can estimate the median, 75th percentile, 90th percentile, or any other quantile. Quantile regression has been around since the 1800s for linear models but is not very well known within the machine learning community. The reasons for this are twofold: there has only been [limited research](https://journals.sagepub.com/doi/abs/10.1177/1471082X18759142?journalCode=smja) on practical applications outside of academia, and all machine learning competitions, such as the popular website [kaggle.com](https://www.kaggle.com/competitions), have focused on model accuracy. \n", "\n", "In [supervised learning](https://en.wikipedia.org/wiki/Supervised_learning), the goal is to model the target conditioned on data. The process of creating a model from data is called *training*. During training, many models are built and compared, but only the best is kept. One of the choices that a modeler makes when fitting a model is what defines a \"best\" model. This is controlled by the loss function. When the value of the loss function is high, then the model is considered \"worse\". When the loss is low, the model is considered \"better\". \n", "\n", "**Theorem 1: Using Mean Squared Error as a loss function results in predicting the mean.**\n", "\n", "This is the default setting in most machine learning software. The proof can be shown rather easily. This mathematical convenience is another reason why MSE is so popular. Before modern computers, models were limited to loss functions which computers could optimize.\n", "\n", "**Theorem 2: Any quantile, such as the median, can be predicted by modifying the loss function.**\n", "\n", "The math behind this is more challenging, but the payoff is greater because predicting quantiles tells more about the distribution than merely the average. \n", "\n", "***Theorem 1 Proof:***\n", "\n", "Let $Y$ by the target value, $\\hat{Y}$ the predicted value, and $n$ the number of observations. The Mean Squared Error (MSE) is the mean of the squares of the residuals. To optimize the model, find the minimum value of the loss function with respect to u.\n", "\n", "$$\\text{Loss} = \\text{MSE} = \\frac{1}{n}\\sum_{1}^{n} (Y_i - u)^2$$\n", "\n", "Take the derivative and set the result equal to zero.\n", "\n", "$$\\frac{d}{du} \\frac{1}{n}\\sum_{1}^{n} (Y_i - u)^2 = \\frac{-2}{n}\\sum_{1}^{n} (Y_i - u) = 0$$\n", "\n", "Because the right hand side is 0, we can remove $\\frac{-2}{n}$. Solve for $u$ to find that the minimum value is at the average.\n", "\n", "$$\\sum_{1}^{n} (Y_i - u) = 0 \\Rightarrow u = \\frac{1}{n}\\sum_{1}^{n} Y_i \\space \\blacksquare$$\n", "\n", "***Theorem 2 Proof:***\n", "\n", "To make inferences about the distribution of $Y$, we first need to define the distribution function $F_Y(y) = Pr(Y \\leq y)$. For $\\tau \\in(0,1)$, the $\\tau$th quantile of Y the number $y_{\\tau}$ such that $F(y_{\\tau}) = \\tau $.\n", "\n", "Define the *loss function* as $\\rho_{\\tau}(y)=y(\\tau-\\mathbb{I}_{(y<0)})$, where $\\mathbb{I}$ is an indicator function. A specific quantile can be found by minimizing the expected loss of $Y-u$ with respect to $u$.\n", "$$\\text{Loss} = E(\\rho_{\\tau}(Y-u)) = (\\tau-1)\\int_{-\\infty}^{u}(y-u)dF_{Y}(y)+\\tau\\int_{u}^{\\infty}(y-u)dF_{Y}(y) $$\n", "\n", "Just as before, we take the derivative with respect to $u$ and set it equal to zero.\n", "\n", "$$0=(1-\\tau)\\int_{-\\infty}^{u}dF_{Y}(y)-\\tau\\int_{u}^{\\infty}dF_{Y}(y)$$\n", "\n", "This equation reduces to\n", "$0=F_{Y}(u)-\\tau$\n", "and then to\n", "$F_{Y}(u)=\\tau$. \n", "\n", "Hence, $u$ is $\\tau$th quantile of the random variable Y. Because the median is just the 50th percentile, setting $\\tau = 0.5$ results in estimating the median. This implies that any quantile can be predicted by setting the corresponding $\\tau$ in the loss function. $\\space \\blacksquare$\n", "\n", "**Intuition:**\n", "\n", "A residual is the error between the prediction and target value, $y_i - \\hat{y_i}$. For RMSE and MAE, positive and negative residuals are penalized equally. This is because both loss functions are *even*. An even function returns the same output for a negative or positive input. \n", "\n", "$$\\text{MSE}(Y - \\hat{Y}) = \\text{MSE}(\\hat{Y} - Y) = \\frac{1}{n}\\sum_{1}^{n} (Y_i - Y_i)^2$$\n", "\n", "$$\\text{MAE}(Y - \\hat{Y}) = \\text{MAE}(\\hat{Y} - Y) = \\frac{1}{n}\\sum_{1}^{n} |Y_i - Y_i|$$\n", "\n", "Intuitively, this means that there the average (or median) residual should be zero. For quantile loss, positive residuals are penalized differently than negative residuals, which causes the model to aim *above* the center (for quantiles above 0.50) and *below* the center (for quantiles below 0.50).\n", "\n", "Say that $Y$ is uniform on the integers 1,..., 10. Then the $dF_Y(y)$ just turns into $\\frac{1}{10}$ since all $Y_i$'s have the same probability, and the integrals turn into summation.\n", "\n", "Then the quantile loss is\n", "\n", "$$\\frac{1}{10}(\\tau - 1)\\sum_{y_iu}(y_i - u)$$\n", "\n", "Ignoring the $\\frac{1}{10}$, this is just\n", "\n", "$$(\\tau - 1) \\sum\\text{(Positive Residuals)} + \\tau \\sum\\text{(Negative Residuals)}$$\n", "\n", "When $\\tau = 0.5$ the positive and negative residuals are weighted equalled. This is just MAE.\n", "\n", "Say that $\\tau = 0.10$ and that the model's \"guess\" for the 10th percentile too low. Then most of the residuals will be positive because the guess is too low. The negative residuals will be very small because they get hit by $0.10$.\n", "\n", "$$(0.1 - 1) \\sum\\text{(Positive Residuals)} + 0$$\n", "\n", "This will result in a small value, or a low loss penalty. This tells the model that a low guess is in the right direction.\n", "\n", "Conversely, say that the model's \"guess\" for the 10th percentile is too high. Then the residuals will often be negative and so the weight will be on the second term and so the loss will be *high*. This tells the model that a high guess is in the wrong direction." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Gradient Boosting\n", "\n", "Gradient Boosting is a technique which forms an ensemble of weak learners to create a single prediction. These weak learners are often decision trees, which means that they can apply to either regression or classification problems. Each tree is built to correct the error of the previous tree. This stage-wise approach allows for hundreds or thousands of decision trees to be ensembled together. A detailed description of the GBM algorithm can be found in chapter 8.2 of [(An Introduction to Statistical Learning)](http://faculty.marshall.usc.edu/gareth-james/ISL/)\n", "\n", "GBMs are an exceptionally versatile method, but they do not apply to all predictive modeling applications. Consideration should be taken to the advantages and disadvantages of the modeling method to determine if it will provide the best solution to the predictive analytics problem.\n", "\n", "**Advantages:**\n", "- High prediction accuracy;\n", "- Has been empirically shown to succeed on a wide class of problems;\n", "- Performs variable selection;\n", "- Handles non-linearities, interaction effects, is resilient to outliers, and corrects for missing values;\n", "- Deals with class imbalance directly by weighting observations.\n", "\n", "**Disadvantages:**\n", "- Requires a large sample size;\n", "- Can takes longer to train than other methods;\n", "- Does not detect linear combinations of features;\n", "- Can overfit if not tuned correctly.\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Application 1: Adding Prediction Intervals to Risk Scores\n", "\n", "-----------\n", "A common use for predictive models is in creating risk scores. A risk score is a value from 1-10, or some other high number, which is intended to measure the tendency of the person to by expensive or not. In health insurance, this is the cost of the person relative to the average. A risk score of 1.5 means that on average, the person costs 50% more than the average person in their population. \n", "\n", "In the table below, three members are represented for the sake of example. Member B at times appears to have the lowest cost, at times when their score is 1.1, but expensive at others, when their score is 10.1. Members A and C have relatively lower variance in their scores. For an insurance practitioner only looking at a single number, at a single point in time, based on a single data sample, all three risk scores appear to have the same precision, but in truth this is not the case.\n", "\n", "![Risk Scores](https://media.giphy.com/media/ME38o1vgzyCIqFY0J5/giphy.gif \"Risk Score Example\")\n", "\n", "The risk score is a random variable. If the data was reshuffled and the model refit, then the risk score will change. A common means of interpreting the average is in terms of frequency. For example, if there were 100 members with the same characteristics as A, if their total health care cost were added together and divided by 100 then this would be about 1.5 times the overall average of A, B, C, and all other members.\n", "\n", "Using quantile regression, we can use GBMs to create a range of values.\n", "\n", "\n", "### Data\n", "\n", "The focus of this paper was on the methodology and reproducibility rather than the result itself. The data consists of 1,338 records of patient health care costs. This data is [publically available](https://www.kaggle.com/mirichoi0218/insurance) for download and was also cleaned and republished as a companion to the textbook *Machine Learning with R* by Brett Lantz. There are six columns which are relate the patient's demographics to their annual health care cost. \n", "\n", "**age**: age of primary beneficiary\n", "\n", "**sex:** insurance contractor gender, female, male\n", "\n", "**bmi:** Body mass index, providing an understanding of body, weights that are relatively high or low relative to height, objective index of body weight (kg / m ^ 2) using the ratio of height to weight, ideally 18.5 to 24.9\n", "\n", "**children:** Number of children covered by health insurance / Number of dependents\n", "\n", "**smoker:** Smoking\n", "\n", "**region:** the beneficiary's residential area in the US, northeast, southeast, southwest, northwest.\n", "\n", "**charges:** Annual health insurance claims\n", "\n" ] }, { "cell_type": "code", "execution_count": 124, "metadata": { "code_folding": [ 15 ] }, "outputs": [], "source": [ "import pandas as pd\n", "import numpy as np\n", "import sklearn as sk\n", "from sklearn import preprocessing, model_selection, ensemble, linear_model\n", "import matplotlib.pyplot as plt\n", "import matplotlib.style\n", "import matplotlib as mpl\n", "import seaborn as sns\n", "import warnings\n", "warnings.filterwarnings('ignore')\n", "\n", "%matplotlib inline\n", "from matplotlib.pylab import rcParams\n", "rcParams['figure.figsize'] = 12, 4\n", "\n", "data = pd.read_csv(\"insurance.csv\")" ] }, { "cell_type": "code", "execution_count": 125, "metadata": { "hideCode": true }, "outputs": [], "source": [ "def style_table(table):\n", " return(table.style.set_table_styles(\n", " [{'selector': 'tr:nth-of-type(odd)',\n", " 'props': [('background', '#eee')]}, \n", " {'selector': 'tr:nth-of-type(even)',\n", " 'props': [('background', 'white')]},\n", " {'selector': 'th',\n", " 'props': [('background', '#606060'), \n", " ('color', 'white'),\n", " ('font-family', 'verdana')]},\n", " {'selector': 'td',\n", " 'props': [('font-family', 'verdana')]},\n", " {'selector': 'th',\n", " 'props':[('max-width', '100px')]}\n", " ]\n", " ).hide_index())" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The top five records are shown below." ] }, { "cell_type": "code", "execution_count": 126, "metadata": { "hideCode": false }, "outputs": [ { "data": { "text/html": [ " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
agesexbmichildrensmokerregioncharges
19female27.90yessouthwest16884.9
18male33.771nosoutheast1725.55
28male333nosoutheast4449.46
33male22.7050nonorthwest21984.5
32male28.880nonorthwest3866.86
" ], "text/plain": [ "" ] }, "execution_count": 126, "metadata": {}, "output_type": "execute_result" } ], "source": [ "style_table(data.head(5))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The the age, sex, bmi, children, smoker, and region columns are used as the predictors. The response variable is the annual claims. One-third of the records are held out as a test set. A random seed is set to ensure reproducibility of results." ] }, { "cell_type": "code", "execution_count": 127, "metadata": {}, "outputs": [], "source": [ "x = data.drop([\"charges\"], axis = 1)\n", "y = data[\"charges\"]\n", "\n", "#convert categories to 0-1 columns\n", "le = preprocessing.LabelEncoder()\n", "x = x.apply(le.fit_transform)\n", "\n", "#train/test split\n", "x_train, x_test, y_train, y_test = model_selection.train_test_split(x, y, test_size = 0.33, random_state = 42)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Categorical columns are converted to binary columns. Two-thirds of the records are used for training and the remaining one-third used as a holdout set. " ] }, { "cell_type": "code", "execution_count": 128, "metadata": { "hideCode": true }, "outputs": [ { "data": { "text/plain": [ "Text(0.02, -0.02, 'The distribution of annual medical claims is right-skewed and has a mean of $13,270.')" ] }, "execution_count": 128, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "font = {'family' : 'normal',\n", " 'size' : 22}\n", "matplotlib.rc('font', **font)\n", "mpl.style.use(\"bmh\")\n", "plt.hist(data['charges'], bins = 30)\n", "plt.gcf().set_size_inches(12, 6)\n", "plt.figtext(.02, -0.02, \"The distribution of annual medical claims is right-skewed and has a mean of $13,270.\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Predicting the Average" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The first step in modeling is to predict the average for each patient. The loss function will be MSE so that the prediction will be the mean, as shown in Theorem 1. \n", "\n", "In addition, a linear model is fit to serve as a baseline against the GBM.\n" ] }, { "cell_type": "code", "execution_count": 129, "metadata": {}, "outputs": [], "source": [ "linear_model = sk.linear_model.LinearRegression()\n", "linear_model.fit(x_train, y_train);" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "A random grid search of parameters is used with ten-fold cross validation. This means that each model trained, the parameters are calibrated based on 90% of the data and evaluated on the remaining 10% (or 1/10th). Once the best parameters have been selected, the model performance is evaluated on the unseen holdout data. Because this is a relatively small data set for a GBM, these measures will help to prevent the model from over-fitting. \n", "\n", "In order to make this example faster to reproduce, only the final parameters are shown. A range of values was tested such as `'n_estimators: [500, 1000, 2000]', 'max_depth':[2,3,4,5]`, and so forth. There are several different methods for tuning parameters in addition to a grid search such as a random search or Bayesian optimization which are beyond the scope of this paper." ] }, { "cell_type": "code", "execution_count": 130, "metadata": {}, "outputs": [], "source": [ "#predict the mean\n", "parameters = {\n", " 'n_estimators': [1000], \n", " 'max_depth': [4],\n", " 'min_samples_split': [5],\n", " 'learning_rate': [0.1], \n", " 'loss': ['ls']\n", "}\n", "mean = ensemble.GradientBoostingRegressor(**parameters, random_state = 42)\n", "mean_cv = model_selection.GridSearchCV(mean, parameters, cv=10)\n", "mean_cv.fit(x_train, y_train);" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The best set of parameters is recorded. " ] }, { "cell_type": "code", "execution_count": 131, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "{'learning_rate': 0.1,\n", " 'loss': 'ls',\n", " 'max_depth': 4,\n", " 'min_samples_split': 5,\n", " 'n_estimators': 1000}" ] }, "execution_count": 131, "metadata": {}, "output_type": "execute_result" } ], "source": [ "mean_cv.best_params_" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Predicting Quantiles" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Recall that any quantile can be estimated. For this example, a symmetric range between the 95th and 5th percentiles is chosen. These bounds will form the prediction interval. Depending on the nature of the problem and the needed level of confidence, these can be made wider or narrower." ] }, { "cell_type": "code", "execution_count": 132, "metadata": {}, "outputs": [], "source": [ "#predict the upper percentile\n", "parameters = {\n", " 'loss': 'quantile',\n", " 'alpha': 0.95,\n", " 'learning_rate': 0.1,\n", " 'n_estimators': 1000,\n", " 'min_samples_split': 5,\n", " 'max_depth': 4\n", "}\n", "\n", "upper = ensemble.GradientBoostingRegressor(**parameters, random_state = 42)\n", "upper.fit(x_train, y_train)\n", "\n", "#predict the lower percentile\n", "parameters = {\n", " 'loss': 'quantile',\n", " 'alpha': 0.05,\n", " 'learning_rate': 0.1,\n", " 'n_estimators': 1000,\n", " 'min_samples_split': 5,\n", " 'max_depth': 4\n", "}\n", "\n", "lower = sk.ensemble.GradientBoostingRegressor(**parameters, random_state = 42)\n", "lower.fit(x_train, y_train);" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "While the parameter values for the `learning_rate`, `n_estimators`, `min_samples`, and `max_depth` are not retuned, technically they should be. As the loss function changes, the minimum value in the parameter space moves. The further away $\\tau$ is from 0.5 the more significant this becomes." ] }, { "cell_type": "code", "execution_count": 133, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "{'r2': 0.75, 'mae': 4242.87}" ] }, "execution_count": 133, "metadata": {}, "output_type": "execute_result" } ], "source": [ "def score_model(model):\n", " scores = model_selection.cross_validate(model, x_test, y_test, cv = 3, scoring = ('r2', 'neg_mean_absolute_error'))\n", " r2 = scores['test_r2'].mean().round(3)\n", " mae = scores['test_neg_mean_absolute_error'].mean().round(2)\n", " return({\"r2\": r2, \"mae\": -mae})\n", " \n", "score_model(linear_model)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The linear model has an R-Squared of 0.75 and an MAE of 4242.9. " ] }, { "cell_type": "code", "execution_count": 134, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "{'r2': 0.809, 'mae': 3160.52}" ] }, "execution_count": 134, "metadata": {}, "output_type": "execute_result" } ], "source": [ "score_model(mean_cv.best_estimator_)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "As expected, the GBM significantly improves on those metrics. The R-Squared is higher and the MAE is lower." ] }, { "cell_type": "code", "execution_count": 135, "metadata": { "hideCode": false }, "outputs": [ { "data": { "text/html": [ " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
AgeGenderClaims5th Prediction PercentileAverage Prediction95th Prediction Percentile
2709095.078689.4310241.722949
1805272.185060.6910599.518030.4
4602933124531.529278.632011.3
" ], "text/plain": [ "" ] }, "execution_count": 135, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df = x_test\\\n", " .assign(charges = y_test,\n", " linear_pred = linear_model.predict(x_test),\n", " pred_lower = lower.predict(x_test),\n", " pred_mean = mean_cv.predict(x_test),\n", " pred_upper = upper.predict(x_test)\n", " )\n", "\n", "p = df[['age', 'sex', 'charges', 'pred_lower', 'pred_mean', 'pred_upper']]\\\n", " .rename(columns = {\n", " 'age': 'Age', \n", " 'sex': 'Gender', \n", " 'charges': 'Claims', \n", " 'pred_lower': '5th Prediction Percentile',\n", " 'pred_mean': 'Average Prediction',\n", " 'pred_upper': '95th Prediction Percentile'})\\\n", " .head(3)\n", "\n", "style_table(p)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The above table shows three members along with a central, middle, and high prediction of their claims." ] }, { "cell_type": "code", "execution_count": 136, "metadata": { "code_folding": [ 6, 9 ], "hideCode": true }, "outputs": [ { "data": { "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "sns.set(font_scale = 2)\n", "sns.set_style(\"ticks\")\n", "\n", "custom_colors = [\"#eb4034\", \"#285bd4\",\"#2bbd35\"]\n", "\n", "plt.figure(figsize=(12, 12))\n", "df1 = df.assign(Range = '5th Percentile').rename(columns={'pred_lower': \"Prediction\"})\n", "df2 = df.assign(Range = 'Mean').rename(columns={'pred_mean': \"Prediction\"})\n", "df3 = df.assign(Range = '95th Percentile').rename(columns={'pred_upper': \"Prediction\"})\n", "long_df = pd.concat([df1, df2, df3])\\\n", " .rename(columns={'charges': \"Claims\"})\n", "p = long_df[['Claims', 'Prediction', 'Range']].where(long_df['Claims'] < 50000)\n", "lm = sns.lmplot(x = 'Claims', y = 'Prediction', hue = 'Range', data = p,\n", " height = 10, aspect = 1.5, ci = None, legend_out = False, lowess = True,\n", " palette = custom_colors)\n", "fig = lm.fig\n", "fig.suptitle(\"Actual vs. Predicted Claims - Holdout Data\")\n", "plt.figtext(.02, -0.1, \"Each patient has one point for their mean, 5th percentile, and 95th percentile prediction. \\nThe lines shown are only a visual aid to show that the 5th percentile is below \\nthe mean and the 95th percentile is above the mean.\");" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "---\n", "\n", "## Application 2: Data Auditing with Outlier Detection\n", "\n", "A colloquial phrase speaking to the importance of data quality in statistics is \"garbage in; garbage out.\" Any result can only be as useful as the data on which it is based. Quantile regression can be used to detect bad data. By training the model on good, useful examples, a rule can be created to detect if a point is an outlier. Out of the 1,300 patients in this study, this method correctly detects only a handful of cases which show irregular values. \n", "\n", "Some of the advantages to this method are:\n", "* Scalability \n", "* No reliance on specialized domain knowledge \n", "* Non-parametric (few statistical assumptions are needed)\n", "* One a model has been trained it can audit new data very quickly\n", "\n", "The [Inter Quartile Range](https://en.wikipedia.org/wiki/Interquartile_range) is a statistical measure of central tendency. This is the difference range between the 75th and 25th percentiles. These are commonly displayed in a box plot as the \"width\" of the box. The center of the box represents the median. \n" ] }, { "cell_type": "code", "execution_count": 187, "metadata": { "hideCode": true, "hideOutput": false }, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAABCUAAAMaCAYAAACxg/xkAAAAAXNSR0IArs4c6QAAAARnQU1BAACxjwv8YQUAAAAJcEhZcwAADsMAAA7DAcdvqGQAAL1oSURBVHhe7N0LfFXVnff/zDMz7ThtHVunT+OtFQtaqm29zFDQtkoqrdCxtbRoDbVCU9rS1KqDaG20IkZqaUSleUQxhosmKBpwokFEA2o0OH81oLRcNUFBI0U9IpdUUX7/vXbWPmefk3Xu+9z2/rxfr99Lz8o5ScjZl7W+Z++1ygQAAAAAAKAACCUAAAAAAEBBEEoAAAAAAICCIJQAAAAAAAAFQSgBAAAAAAAKglACAAAAAAAUBKEEAAAAAAAoCEIJAAAAAABQEIQSAAAAAACgIAglAAAAAABAQRBKAAAAAACAgiCUAAAAAAAABUEoAQBACVo3vUzGlcWrsbJ8m36iwfvbOmX5zGq5fORQqQy/ZpBMGlUltzS2SfdO/cQ4Ev7s8qFSPWqszJjdJttD+gUAAABxEEoAAFCCMgoldq+X5ZdWuIKIeDVILp/ZKfEyhcQ/21XlFVLX2q1fBQAAMBChBAAApWxbk9ToEKCmuVc3GuzulIbh7sBgqNRMrZfly9qla6VVSxpkzuQxUYFF5QVNskO/3C0SSlTInMX69braG+vklnGDwt9jXNkIaVjdp18JAAAQjVACAIBSllIo0S3LLygPBwWTLmqR7fv1l2Lt7JSFZ7ueO71T9ugvOSKhRPzbRLYvrIwEHKMbZLtuBwAAcCOUAACglKUQSoRaq8MBQbyrH6JEXVUxRlq26HYtlVBCBSGto53nVUsH80sAAAADQgkAAEpZ0lDCHQ5USXuSSSwd76+ulQn6+1Ze2S7v63YltVAi9ecBAIDgIpQAAKCUJQslXF8fNzU6XEisW1pG6deV18o63aqkFjb0yTNTnedNkWeYVgIAABgQSgAAUMqShBLvr5iigwHr64vjzTlhtm6GM7fEWGnt0Y2WlEKJvna5pVw/b1wKt4wAAIBAIpQAAKCUJQkldjSP1QFCmTSs1o0pivfa5KFESJ65coR+TrncsoLLJAAAgBmhBAAApSxJKLFp1klJAoT43FdZmEOJ2CVBW6Rl5hSpOcG5wqJcqmcMXL0DAADAQSgBAEApK2gokaAGj5FblnTrVwAAAJgRSgAAUMoKevtGvBq4jCgAAIAJoQQAwFdmzZplV2CkMdHljNaQbk1NuhNd7lldK5P0zxo3vFbW7dZfCLDAbY8AAKSJUAIA4CtHHHGEDBkyRD8KgCShhPvruV8StM/6mjPBZZlUXsCqG8OGDZODDjpIPwIAALEIJQAAvkIoEatbWkc7IUKVtO/UzUm8v7pWJjjhwpXRYUb8UELpluUXOFdYlMmk6cGe6JJQAgCAxAglAAC+QigxUKi1Wir1c1K6emF3pzQMd4KHgfNDJA4lLFGvL7d+r+BOeEkoAQBAYoQSAABfIZQwibl64aIm6Y43vURvuzSc7XrujK4Bt3wkDSWUHuv3KneeN0IaVvfpLwQLoQQAAIkRSgAAfIVQIo6oqxesKh8qNVPrZfmydulaadWSBpkzeUz4igpVlRNbjFdVpBRKWKImviyvlOWuyTKDglACAIDECCUAAL4SlFAiEgyYKk5YEFovrRMHGZ4fW4Pk8tldceeCSDWUUHY0V0aCjgCuyEEoAQBAYoQSAABfIZRQleQKhi3tsnxmtVw+cmjUlRF2TWqSHUnutEgnlBDpjbp1ZMLVwZr4klACAIDECCUAAL4SuNs3PNElC0/UQcM4lvH0EqEEAACJEUoAAHyFUCITffLMVOfqh4GrbSBzhBIAACRGKAEA8BVCicy8v7pWJuhbLMaVV8iM2S32BJgdzfVyy/lDE0+iibgIJQAASIxQAgDgK4QSmeqTTbOiV99wF6FEZgglAABIjFACAOArhBLZCT3bJHMmVsiEch1IlA+V6lFVsnB1SD8D6SCUAAAgMUIJAICvEEqgmBBKAACQGKEEAMBXCCVQTAglAABIjFACAOArhBIoJoQSAAAkRigBAPAVQgkUE0IJAAASI5QAAPgKoQSKCaEEAACJEUoAAHxFhRJHH320dHd3U1TB6ytf+QqhBAAACRBKAAB8RYUSH/3oR6WsrIyiCl6f+tSnCCUAAEiAUAIA4CsqlDj44INl4sSJFFXw+vd//3dCCQAAEiCUAAD4CnNKoJgwpwQAAIkRSgAAfIVQAsWEUAIAgMQIJQAAvkIogWJCKAEAQGKEEgAAXyGUQDEhlAAAIDFCCQCArxBKoJgQSgAAkBihBADAVwglUEwIJQAASIxQAgDgK4QSKCaEEgAAJEYoAQDwFUIJFBNCCQAAEiOUAAD4CqEEigmhBAAAiRFKAAB8hVACxYRQAgCAxAglAAC+QiiBYkIoAQBAYoQSAABfIZRAMSGUAAAgMUIJAICvEEqgmBBKAACQGKEEAMBXCCVQTAglAABIjFACAOArhBIoJoQSAAAkRigBAPAVQgkUE0IJAAASI5QAAPjK2rVr7QKKAdsjAACJEUoAAAAAAICCIJQAAAAAAAAFQSgBAAAAAAAKglACAAAAAAAUBKEEAAAAAAAoCEIJAAAAAABQEIQSAAAAAACgIAglAAAAAABAQRBKAAAAAACAgiCUAAAAAAAABUEoAQAAAAAACoJQAgAAAAAAFAShBAAAAAAAKAhCCQAAAAAAUBCEEgAAAAAAoCAIJQAAAAAAQEEQSgAAAAAAgIIglAAAAAAAAAVBKAEAAAAAAAqCUAIAAAAAABQEoQQAAAAAACgIQgkAAAAAAFAQhBIAAAAAAKAgCCUAAAAAAEBBEEoAAAAAAICCIJQAAAAAAAAFQSgBAAAAAAAKglACAFC0QhvapGVqlVQPHyTjysr6q3yoVI+rloXNnbKjTz8xZSHpmNz/fRpW6yYgRVlvj/tD0r2ySRbF+R6LlnXLHv1UAACCglACAFB8dnbKwnGuQVu8Kq+Quub1qQ/kdrfLLeX9ryWUQMq82B57W+Rave0lqsqRtdIV0q8BACAACCUAAMWlp0lq3IO3wWNkxtUN0r6yXbqs6miul1vOHyqVroHcpOmdiYOJ/X2yY0ObNJxdHn4NoQRS4tX2uM36PvrrlcOr5JbZTdJhf482WT57ilw+OPL6ygtaZId+GQAAfkcoAQAoHrs7pWG4Mzgrl8tndcUNG97f0iJ1rufWNHfrr7j1yvJxznOii1ACSXm5PW6zvj6pXrq2xbnHI+pnjZXl23Q7AAA+RygBACgam2aN0IOyFK5+UNyfYpdPkWd26/awXmm/tEKqR+kaGflEm1ACyXi/PSa2Y+HY8M9j+wQABAWhBACgOPRF5nsYd2KdbNLNyexojgzkLmk0XS3h4rqEnkEfEsrH9hgjEkpwpQQAIDgIJQAAxWF1bfgqhuq5aQzm1OBRv27cuKbE9+ITSiBV+dgeo3RLy+j+11VObhPmugQABAWhBACgKGyfW9E/kCsrl4ZndWNKemX5BXoQWFYjXbrViFACKcrL9qjs75Ptq5tkjp6EVa2+sS7N2z4AAChlhBIAgKKwbrozkEv/0vWUX0sogRTlent03+ahqvKESpnT3Ck79usnAAAQEIQSAIAiEJKOyc4ArVbW6dZURT7VJpSAF3K/PcaGEuqKjAnn18ryZ7lxAwAQLIQSAIAiQCiBYpKH7bEvJHtCqrpl08oWWThukH5NvOVtAQDwJ0IJAEBR4PYNFJO8bI9R+qzXOUuQjpHWLboZAACfI5QAABSFyKfLJ8nCNboxJUx0Ce/lZXuM1dMkl+vtM60VPwAAKGGEEgCA4sCSoCgmeV8S1LK7Teqc107v1I0AAPgboQQAoDiowVy5HpCdWCebdHMy7gkDL2lMMngklECq8rE9xnJtn4QSAICgIJQAABSNTbOce+rLZJI1KNuj2+PqsQZxzsCxfIo8s1u3x0MogTTkfHuMEVpSFf55NYt7dSsAAP5GKAEAKB67O6VhuB7UlZXL5TPaZcd+/bUYe9Y1SZ3ruSkN4gglkA4Pt8d1M8fKwmXrJRTn9dGBRrV07NTtAAD4HKEEAKC4uAdnqgaPkRlXN0j7ynbpsqqjuV5uOX9o+H5/NQCsnpHCp9gKoQTS5dH2GF6Ro3yoXDK5RhY1t9mv71rWJAsnj4l6PUuCAgCChFACAFB8ettkzshyPUhLUOUVUrck0QCuV5aPM7xuQNXKOv0KYAAPtsdNMweZXxNVg+TaxvWpBWwAAPgEoQQAoGjtWNMiLVOrpHr4wAFd9czO+JfChxFKwDvZbo/v93RJe2ONzBhXIZMGR15bOXyszJjZJOuYRgIAEECEEgCA0hFqkxnOQI7VCVBobI8AAGSNUAIAUEK6pWWU/oQ5g9UNAG+xPQIAkC1CCQBASdnRPLZ/EKhq8FiZoycMtC+LHz2ICSyRV2yPAABkh1ACAFBiuqXjohGRgWBMMQhEfrE9AgCQDUIJAEAJ6pPty+plxrgRkaUUB4+Q6nHV0r5FPwXIG7ZHAAAyRSgBAAAAAAAKglACAAAAAAAUBKEEAAAAAAAoCEIJAAAAAABQEIQSAAAAAACgIAglAAAAAABAQRBKAAB8ZVpZGUUVXQEAADPOkgAAX1EDwPlnnEFlWDM+/nG5/mMfM9bc//gP42uoxEUoAQBAfJwlAQC+4oQSyMycr3wl/Ol+bPWuWaOfhVQRSgAAkBhnSQCAr6gBIKFE5gglvEUoAQBAYpwlAQC+ogaAhBKZI5TwFqEEAACJcZYEAPiKGgASSmSOUMJbhBIAACTGWRIA4CtqAEgokTlCCW8RSgAAkBhnSQCAr6gBIKFE5gglvEUoAQBAYpwlAQC+ogaAhBKZI5TwFqEEAACJcZYEAPiKGgASSmSOUMJbhBIAACTGWRIA4CtqAEgokTlCCW8RSgAAkBhnSQCAr6gBIKFE5gglvEUoAQBAYpwlAQC+ogaAhBKZI5TwFqEEAACJcZYEAPiKGgASSmSOUMJbhBIAACTGWRIA4CtqAEgokTlCCW8RSgAAkBhnSQCAr6gBIKFE5gglvEUoAQBAYpwlAQC+ogaAhBKZI5TwFqEEAACJcZbMh75eWddcJzPGVciE8jIZZ3VOVFUOr5BrpzZIx4aQfmIa1tTJBPV9xjXJDt1U+rpk4YnqbzNWlm/TTRnrlAb9d05eXvy8XOuV5eNMv7tTtbJOP9MktKFNWqZWSfXwQZHXlA+V6nHVsrC5U3b06ScaJf5bVp5QYX+fRcu6ZY9+BVBIagBIKJE5QglvEUogKx70Id/f1iXtjTXm7zG9Sdb16ieWNPqQ8WXRh9wfku5lDXLLxAqZNDjyGtX3u3xynSxf3Svv66ea0YdEajhL5lSfdDdPkWrXCSBeVY6rk66d+mUp2D53TP9r/RRKbGmQS+y/ByeUgTI8oezslIXjXEFEvCqvkLrm9XFOCKn/LStHTpGOHv0yoEDUAJBQInOEEt4ilEBmvOlDbpqZQh+gbJBc29ytX1Gi6EMmkFkfMtRRJ5e7goh4pfp+yzfE+3SLPiRSw1kyZ/pk3fQRrp2tXCacP0UWNbdJ18p2q1qk5epqucS9s5dXyvIkO+P7oW5Z11gtk5zX+CGU6AvJ9tUNUjfc+Vt4fEKZVK//5vGqK8mVAkVmda3+O5VJw2rdZtLTJDXuzszgMTLj6gZp1//ujuZ6ueX8oVLpfN2qSdM7DcGE6285qib8+v6ytuOZU6JPWsOtk9tu/VKgANQAkFAic4QS3iKUQPq860Oum+58fahcMrlGWpbo8/eSBpkzqcLVBxghDc+WUmdIow+ZnhT7kDuaK6P6h5Wjq2VOY4v+N7fJ8tlTpOaE8vDX7e1ntekPQR8SqeEsmSNRO/PwammPFzbsD8m6WWNcz42zM7oOIlFV0qFEvPTU4xOKNdD2lVROKLutf3/4BF0ul8/qintZ3PtbWlwn83KpGfBpietvGW97298trRdETk6XNJb4Jy4oaWoASCiROUIJbxFKIF1e9iHXzRwrc5Z0SWi/bojh/lmVHveXdjSP7f+9ctJXpQ+ZkRT6kHus54Q//CwfIwufjXeLUJ9sX+L6oNQYjNGHRGo4S+bCzjaZ4XxCncLVD7GJ+CVzDTvjs/VSPaoiXOH7uko6lOiSRa5/U2S+g9I5oexYXJfC++voluUzW7J/v1I4oWyaFdmezFc/xHBfVVE+RZ6J6tSkcEJRwpdOWjW5jXsDUTBqAEgokTlCCW8RSiAtuehDJtQpc/Rrve5T5jaUoA+ZkaR9yC5ZGP6gKt7VD9Gigq2p7TH9P/qQSA1nyRzY3qjne7Dq2iUpTmLZZ+209gQ9VpXXyDNJjgHhy/FKOpSIET5QlsYJJdRa1X8QTqnTYJ1MJvafMCsntkgGU5tGJDuh9LXLLU6H5sQ62aSbkwl3HqyKTqlTPKGk/Dwgt9QAkFAic4QS3iKUQDry0YeMFgklSutKiRj0IVOTpA/5/sqacMAwYWaXbk3GPWfFGGndoptt9CGRGs6SnnPvmFPSOjFsn1uhX1ce9xNwB6FEMrk/odgnCedys4QnlcjJRD2vNeVUPI5koYT1deeEUp3OJyYqzNCvi96uUjxRuF8/tT3JbMxA7qgBIKFE5gglvEUogdTlpw/ptmfFFN1nGCELPd69CSUSKc4+5Lrpzm0UFdISFS4k9r61HTnft6bZvZwLfUikhrOk19w71QVpppnPRgaTly9MPJgklEjGdRCMKW+XwEp2UvH4ZKIkOaFEdUye1Y0psTpDF/R/33FlNRLJx1M7objT9egTEpBfagBIKJE5QglvEUogZXnqQyrv71wvHbOr9HwAg+Taxene9pGcL0KJmPJ3H7JbWkbpf2u5eUWOuHpbpEZ/33FXu4Mc+pBIDWdJr21riuyU6aarabyWUCKZ+CeUSHm1BFb0SSNyUumWjov0fZ5enUyUJKFEeNvI4O9ofm0KJ5SdVkfKuQdxwJwUQH6pASChROYIJbxFKIGU5bwPGds3GiSXTK6Xji1Z3RAQlx9DiUj5sQ+ZWoBgFu+19CGRGs6SXltTJxP0zpd20udOyAklstQpC4dXyS1q+aI13bInFLJr+5p2aZ9ZKRPCS2WWyy0r0rr5Mg7XSWVwlXVSiT6ZxL8sLwMJTygh6Zjs/NvSTLktkass4oQSscs5LWmQhqmuv2d5hfU7efH3BDKnBoCEEpkjlPAWoQRSlvM+pGGwrZYLn90m23MwECzlUCKQfchQm8zQX0s7FLP+DeGrLOKFEvQhkQBnSa+VaCgR+YQ8g0r7wBWHpyeUxPZsqI98opDGZJCJuU4qTnl9MlEKGUokqMrRU+IvWwbkkRoAEkpkjlDCW4QSSFke+pDv6wH2jg2d0tE4RaqdAWG8JemTcfVJMir6kFoR9CFzHUokKPqQ4CzptRK9fSNooYQSmcxnrHeXxe23Dr5fcP42Qw2hgQcShhLu9zL9v6P5tSmcUM5ukO362UChqQEgoUTmCCW8RSiBlOWpDxnFtSR45ZUZTDAYwFBC8Wcf0tXfS3t8Ee+19CGRGs6SXnMn1emus1vIiS539yfnGZVXl/zl+YQSvqzQKm8O/MVwpYT7aoeT0pxJO92JLqPXRp9kdSxYVxrFQA0ACSUyRyjhLUIJpCxPfchYXVfrn5nmih+2/X3mvqFV3XN1P+v7DdJt+Lpd9CG1YuhDuq52SPcKkLQnuqQPiWicJT3HkqAZK+kTiutkYp9EYh/rp3khSSihvu50THK+JOhu62vO5ETWdlvjyaRPQHbUAJBQInOEEt4ilEDq8r8kqNJ9qzM49Lb/VbpzSiTn1z5kXpcEpQ8JF86SObC9cUx4x7x2SYozGvdZO+aJescsr0l6IiKUyN6mWSeFf152l97FO3lEt+dr9Q07XHDuEU0j6XafYC9pdJ8YEpxQFNeln2qdcyYpQqGpASChROYIJbxFKIF05KMPGSub2z4T8XMo4dc+pHtpzgkzI9fMJuYO08ZIa1SYQR8SqeEsmQs722SGs4OllHBGX8J0SQqfbhNKZKtLFjon8KwmKUqWZufgpJIslLBsmpXmJXHuk8KA5ZiSnFAse6zfqX+tc/V6j1N9IE1qAEgokTlCCW8RSiAteehDRnP1h8rTnyA7Ef+GEn7uQ1r/tvDVC6mFBDuaK8NBRuXU9pg+J31IpIazZI64d9Bxw6tl+bo4aff+XumaMcb1XOuEkEJIGPhQwnVQHfg36JXll1bJopW9cSZsyvYE7rBOFhfoy9wSHkQ9PqkkPaFYYi6Ju3xGu+zYr78WY8+6JqlzXz63OHbG7+QnFCV6m89wFm/AA2oASCiROUIJbxFKIF2e9SG3tUjdpQ2yblu8jmVIutz39c9K9ZPx1BRtKEEf0q54fcjokGCMNFh/C7OQdDdXRwcKA/729CGRGs6SORN90FKDvQnnT5FFzW16fd4Wabm6Wi4Z7HzdKnttYv3yJAglkpxQnMvIyofKJZNrpGVJ/5rIHc11MmO0PrirSjEEMgktqeo/gKZ0koicfContliH8SykcEKxRV0SZ5Vai/zqhvAa0R3N9XLL+UMjJwFrG62eYbqqIrUTSuw2X3mBj7ZNlBQ1ACSUyByhhLcIJZA+j/qQrhU5KodXyS0zI32A9sYpUnOCM3+AVTkYCJZsKEEfMjokUD93dLXMaWzR21+bLJ8ds/2UV8S5qoI+JFLDWTKnQtLdqA86SapyXL3EC8Jt7gNogkp7XeuCch2oEpXpIJbshBJeRSJ+VY6sla6sjuzWQXtxXRqXmVknlZlNsj3OFQuJuU6SxrJOjPqZUXrbZM5I10kjXlknk7ol8dL+VE8oiiv5t4rZlFEIagBIKJE5QglvEUogMx70IXtb5Fr3hxNxqnJkjTyzU7/GQ7kNJehDpi6zPuSOZTVSndL2M0Xa4/476EMiNZwl8yHULV0qXR1XIRMG7NxVsjzuZXUuhBLREp5QlD7ZsaZFWqZWSfVI19UA5UOleuIUaVnWXWIHugxDCS38txjuSvh1Vc/slFDCk1w6JxTLthbXFRonyZyODD9GADKkBoCEEpkjlPAWoQSykm0fsi8k3SubZFEh+kPWz+5f+jMX/QD6kKnLog+5v1c2LWmQWyZWyCT3lTl2VUhDR7Jkhj4kUsNZskDcs/YmvcwM8FqoTWbok0TldPd60kDpUwNAQonMEUp4i1ACXqMPiULa01qtt7/0l58F4uEsWSDu9XwznyQHyFS3tIzq3/4GrrYBlDY1ACSUyByhhLcIJeA1+pAoqC0NUq23v4GrbQCZ4SxZKO41pdUEg5Pq+ycfWqYusauUCUzwghwL3+upavBYmaMn0GpvrLEnciL9RqlSA0BCicwRSniLUAKeow+Jgoq+HaRyXK0sX6YmwHQmYE18SzFgwlmygPasqZPL400gk8p9V0BWuqXjIvfs3tFFKIFSpQaAhBKZI5TwFqEEcoE+JAqqp821nHxsEUogfZwlC21nlyyfXiXV4WV1ymXCyAq5dhYzziIf+mT7snqZMW5EZCKnwSOkely1tG/RTwFKjBoAEkpkjlDCW4QSyBn6kCik3d3SMbtaLndNol45vEIun9wi2/VTgFRxlgQA+IoaABJKZI5QwluEEgAAJMZZEgDgK2oASCiROUIJbxFKAACQGGdJAICvqAEgoUTmCCW8RSgBAEBinCUBAL6iBoCEEpkjlPAWoQQAAIlxlgQA+IoaABJKZI5QwluEEgAAJMZZEgDgK2oASCiROUIJbxFKAACQGGdJAICvqAEgoUTmCCW8RSgBAEBinCUBAL6iBoCEEpkjlPAWoQQAAIlxlgQA+IoaABJKZI5QwluEEgAAJMZZEgDgK2oASCiROUIJbxFKAACQGGdJAICvqAEgoUTmCCW8RSgBAEBinCUBAL6iBoCEEpkjlPAWoQQAAIlxlgQA+IoaABJKZI5QwluEEgAAJMZZEoG1a9u2cAHwDzUAJJTIHKGEtwglAP+hDwl4i7MkAmtTa6vdUeysq9MtAPxA7deEEpkjlPAWoQTgP1133mnv12sXLNAtALLBWRKBRSgB+JParwklMkco4S1CCcB/CCUAb3GWRGARSgD+pPZrQonMEUp4i1AC8B9CCcBbnCURWIQSgD+p/ZpQInOEEt4ilAD8h1AC8BZnSQQWoQTgT2q/JpTIHKGEtwglAP8hlAC8xVkSgUUoAfiT2q8JJTJHKOEtQgnAfwglAG9xlkRgEUoA/qT2a0KJzBFKeItQAvAfQgnAW5wlEViEEoA/qf2aUCJzhBLeIpQA/IdQAvAWZ0kEFqEE4E9qvyaUyByhhLcIJQD/IZQAvMVZEoFFKAH4k9qvCSUyRyjhLUIJwH8IJQBvcZZEYBFKAP6k9mtCicwRSniLUALwH0IJwFucJRFYhBKAP6n9mlAic4QS3iKUAPyHUALwFmdJBBahBOBPar8mlMgcoYS3CCUA/yGUALzFWRKBRSgB+JParwklMkco4S1CCcB/CCUAb3GWRGARSgD+pPZrQonMEUp4i1AC8B9CCcBbnCURWIQSgD+p/ZpQInOEEt4ilAD8h1AC8BZnSQQWoQTgT2q/JpTIHKGEtwglAP8hlAC8xVkSgeWEEnO+/GVZ/MMfUhTlk1L7NaFE5gglvOWEEqZtlaKo0qy5//Ef9n5NKAF4g1ACgeWEEhRF+a8IJTJHKOEtJ5SgKMp/RSgBeINQAoHlhBKPX3ut7N25k6Ion5TarwklMkco4S0nlDBtqxRFlWY9c8st9n5NKAF4g1ACgcWcEoA/qf2aUCJzhBLeckIJAP7BnBKAtzhLIrAIJQB/Uvs1oUTmCCW8RSgB+A+hBOAtzpIILEIJwJ/Ufk0okTlCCW8RSgD+QygBeIuzJAKLUALwJ7VfE0pkjlDCW4QSgP8QSgDe4iyJwCKUAPxJ7deEEpkjlPAWoQTgP4QSgLc4SyKwCCUAf1L7NaFE5gglvEUoAfgPoQTgLc6SCCxCCcCf1H5NKJE5QglvEUoA/kMoAXiLsyQCi1AC8Ce1XxNKZI5QwluEEoD/EEoA3uIsicAilAD8Se3XhBKZI5TwFqEE4D+EEoC3OEsisAglAH9S+zWhROYIJbxFKAH4D6EE4C3OkggsQgnAn9R+TSiROUIJbxFKAP5DKAF4i7MkAotQAvAntV8TSmSOUMJbhBKA/xBKAN7iLInAIpQA/Ent14QSmSOU8BahBOA/hBKAtzhLIrAIJQB/Uvs1oUTmCCW8RSgB+A+hBOAtzpIILEIJwJ/Ufk0okTlCCW8RSgD+QygBeIuzJAKLUALwJ7VfE0pkjlDCW4QSgP8QSgDe4iyJwFKhxMqrriKUAHxGdRQJJTJHKOEtQgnAf1QoofqQhBKANzhLAgB8RQ0ACSUyRyjhLUIJAAAS4ywJAPAVNQAklMgcoYS3CCUAAEiMsyQAwFfUAJBQInOEEt4ilAAAIDHOkgAAX1EDQEKJzBFKeItQAgCAxDhLAgB8Y+PGjfYA8Jf/8i+ydetW3Yp0EEp4i1ACAIDEOEsisLZs2SIzZsyQ//3f/9UtAPxADQC5UiJzhBLeuuELX7D/dgD844UXXrD7kC+++KJuAZANzpIIrNbWVimzOop1LAkK+IoaABJKZI5QwlsXfeIT9t8OgH80NjbafcgFLAkKeIKzJALrb3/7mzzyyCPS09OjWwD4gRoAEkpkjlDCW7NPPtn+2wHwj+3bt9t9yNdee023AMgGZ0kAgG8wp0T2CCW8xZwSAAAkxlkSAOAboVDIHgD+cehQ2b17t25FOgglvEUoAQBAYpwlEVhvv/22dHZ2cukd4DNqAMjtG5kjlPBW/X/+p/23A+AfO3bssPuQ6lZgANnjLInAYqJLwJ/UAJBQInOEEt5iokvAf5joEvAWZ0kE1oYNG+TKK6+Ujo4O3QKg1KlPr+wB9NFH21dDIX2EEt66fsgQ+28HwD+ee+45uw/5/PPP6xYA2eAsCQDwDSa6zB6hhLeYUwIAgMQ4SwIAfEUNALl9I3OEEt4ilAAAIDHOkgAAX1EDQEKJzBFKeItQAgCAxDhLIrAeeugh+chHPiI33XSTbgHgB2oASCiROUIJb138b/9m/+0A+Mf8+fPtPuRdd92lWwBkg7MkAuupp56SU089VZqbm3ULgFLHnBLZI5Tw1pWHHWb/7QD4h/pgS/Uhly1bplsAZIOzJADAN3bu3GkPAGsHD5Z33nlHtyIdhBLe4vYNAAAS4ywJAPAVNQDk9o3MEUp4i1ACAIDEOEsCAHxFDQAJJTJHKOEtQgkAABLjLInAWr16tYwaNUruv/9+3QKg1L311lv2AHDGscfKrl27dCvSQSjhrZojjrD/dgD8Y/ny5XYfcsWKFboFQDY4SyKwWltbpczqKNbV1ekWAKWOiS6zRyjhrYs+8Qn7bwfAPxobG+0+5IIFC3QLgGxwlkRgffjhh/Lee+/Z/wXgH2oAyO0bmSOU8Na800+3/3YA/IM+JOAtzpIAAF9RA0BCicwRSniLOSUAAEiMsyQAwFfUAJBQInOEEt4ilAAAIDHOkgiszZs3y/XXXy/PPPOMbgFQ6phTInuEEt664QtfsP92APxj7dq1dh/yxRdf1C0AssFZEoHFRJeA//T29toDwN9/9rPy5ptv6lakg1DCW0x0CfgPE10C3uIsicD629/+Zi/lxKepgL+oASC3b2SOUMJbs08+2f7bAfCP1157ze5Dqv8CyB5nSQCAr6gBIKFE5gglvMWcEgAAJMZZEgDgG7t27bIHgHUnnCB79uzRrUgHoYS3CCUAAEiMsyQC66233pKnnnqKS+8AH2Giy+wRSnir/j/+w/7bAfCPHTt22H1I9V8A2eMsicBiokvAn9QAkNs3Mkco4S0mugT8h4kuAW9xlkRgbdiwQWpqauykG4B/qAEgoUTmCCW8NePYY+2/HQD/eO655+w+ZFdXl24BkA3OkgAAX1EDQEKJzBFKeIs5JQAASIyzJADANzZt2mQPAH/1r/8qr7zyim5FOgglvEUoAQBAYpwlAQC+sW3bNnsA+NvyciYgyxChhLcIJQAASIyzJALroYcekn/6p3+SWbNm6RYAfqAGgNy+kTlCCW/95uCD7b8dAP+YN2+e3YdcuHChbgGQDc6SCCw1weVpp50mixYt0i0A/EANAAklMkco4a0rDzvM/tsB8A/1wZbqQy5btky3AMgGZ0kAgG/s3bvXHgDOPvlk6evr061IB6GEt7h9AwCAxDhLAgB8Y+PGjfYA8Jf/8i+ydetW3Yp0EEp4i1ACAIDEOEsCAHxj//799gDwjtNOkw8++EC3Ih2EEt4ilAAAIDHOkgis1atXy5lnnin33XefbgHgB2oAyJwSmSOU8FbNEUfYfzsA/vHwww/bfcgVK1boFgDZ4CyJwGptbZUyq6NYV1enWwD4gRoAEkpkjlDCWxd94hP23w6AfzQ2Ntp9yAULFugWANngLInAOnDggH2p94cffqhbAJS6l19+2R4AXvxv/ybbt2/XrUgHoYS35p1+uv23A+Afqu+o+pCqLwkge5wlAQC+0dPTYw8Apxx6qLz++uu6FekglPAWc0oAAJAYZ0kAgK+oASC3b2SOUMJbhBIAACTGWRKBtXnzZrnuuuvsCS8B+IcaABJKZI5Qwlt/OO44+28HwD/Wrl1r9yFfeOEF3QIgG5wlEVhMdAn4z3vvvWcPAOcMHy7vv/++bkU6CCW8xUSXgP8w0SXgLc6SCKydO3fKY489Jq+88opuAVDqNm7caA8Af/kv/yJbt27VrUgHoYS3/nzKKfbfDoB/vPbaa3YfkrmLAG9wlgQA+EZfX589APx///mf9lUTSB+hhLeYUwIAgMQ4SwIAfEUNAJlTInOEEt4ilAAAIDHOkgist956Szo6OmT79u26BYAfqAEgoUTmCCW8Vf8f/2H/7QD4xxtvvGH3IXfs2KFbAGSDsyQCi4kuAf9Rc8SoAeDUT39aHrrsMlk1bRqVZv3pM5+JCiLc9eAvf2l8DRW/mOgS8B8mugS8xVkSgaUmxLvqqqvkqaee0i0ASt2WLVvk3P/7fwcMpimq0AXAP55//nm7D7mGq8cAT3CWBAD4Ts+qVRRVVAUAAMwIJQAAAAAAQEEQSgAAfOPDDz+UPXv2yN69e+XAgQO6FQAAAMWKUAKB9eCDD8o//uM/yqxZs3QLgFKn5oo5/PDD5ZhjjpGtW7fqVqBwhg8fLh//+Mf1IwB+MG/ePLsPuXDhQt0CIBuEEgisp59+Wr7+9a/LPffco1sAlLrdu3fLihUr5LHHHpN9+/bpVqBwfv7zn8uZZ56pHwHwg7a2NrsP+fDDD+sWANkglAAAAAAAAAVBKAEAAAAAAAqCUAIA4Buvvfaa/OIXv5Bf//rXsnPnTt0KAACAYkUogcDq7OyUiooKWbx4sW4BUOqY6BLF5le/+pWMHj1aPwLgB2ouCdWHfOSRR3QLgGwQSiCwWltbpaysTOrq6nQLAADeGjZsmBx00EH6EQA/aGxstPuQCxYs0C0AskEogcA6cOCAfPjhh/Z/AQDIBXWeUQXAP+hDAt4ilAAAAAAAAAVBKAEA8A3mlAAAACgthBIIrE2bNsn06dNl9erVugVAqQuFQrJw4UJpbm6W3bt361agcBoaGmTGjBn6EQA/WLNmjd2HfOGFF3QLgGwQSiCwmOgSAJBrTHQJ+A8TXQLeIpRAYO3cuVPa29vllVde0S0AAHjr2Weflccff1w/AuAHr7/+ut2H7O3t1S0AskEoAQDwjR07dsjvf/97ue666+Ttt9/WrQAAAChWhBIAAN9goksAAIDSQiiBwHrzzTfliSeekG3btukWAAC81dXVJR0dHfoRAD9Qt22oPuQbb7yhWwBkg1ACgcVElwCAXGOiS8B/mOgS8BahBAJLXeat7j1/+umndQsABMi2JqmxOtXjrKppZrK2XLn99tvtOU4A+Ie6Akr1IdXSoACyRygBAPCNQs0psaN5rD24d+raJSH9lUS6ZOGJkdeMK6uVdforeRE3lOiUhkL9TgAAIHAIJQAAvqGW+p05c6bcdNNNEgqlEgx4IzaUGHdinWzSX4sntKQq+jWEEgAAIIAIJQAAyNKAUKKsXGa0JgpFuqV1dP/zKsud1xRLKNEtHTPrpMGuNtmuWwEAAHKBUAKB9eCDD8o//MM/yI033qhbACAzkVBirNxy5Zj+/x/dEHdA//7KGqnUz2m5uj8YKJ5QAl766le/Kh/72Mf0IwB+oCa6VH1IJroEvEEogcBSE1x+4xvfkHvvvVe3ACh1b731ltx6660yd+5c2bVrl27NPXcosXxLu9xiX/1QLres7NPPcAtJ+8T+MEDNPbFuev//E0r40y9+8QsZNWqUfgTAD9ra2uw+5PLly3ULgGwQSgAAfKPwE12OleXbRDbNGtH/2HS1xJo6maS+puedSCWUeH9bu7RMrZLqE8r1c8tlwsixMmN2m2xPOHVGn+xYWS8zxo3ovzLDqsrh1utmtsimNQ1xQ4nw7zSuSXboNltfr3Q118ucyWOleuTQ8Pfs/74VcvnkeunYYv6FBvw7Q93SMbtaLnH+TYNHyOVTG2TdTvvpAAAgIAglAADIUmwoITtb5Fr78UnSsNp9tUSfPHNl/yDcCQIShxJ9smnWmKjB/4AaXCXLe/TT3Xavl9aJg8yviamUQwnX1RXxq1wun9Ule/RLHJF/Z420r6iR6vBcGjFVXmn+9wAAAF8ilAAAIEsDQgnpswbhJ/W3TWyR8LUDzqC+vEae0VlF/FDC+h4z9BUXZSOkrrFTtu/WX7K+FlrTJHUj9VUGw63XRt0p0i3LL3CuqiiTypFVsrC5TbpWtkvXsiZZNLVSJrhCgXRCiWuHV8ktMxukfWWndPeGZE9IVbdssr7vnLOdn3mSLIxZvj/y79RVPlRqpqrv0/87LZxUEQlf3H8zAADga4QSCKzOzk4ZOXKkLF68WLcAQGYGhhKWLQ1yid0WGaBvmtkfVFwyt7u/wRI3lHBu8ygbEXO1hcvONpmhwwU1P4Uj1FodHuBPmt454KoFW0+TXK6fk3IokUxfpzSc2P/ayhldurGfO5SYdFGTdA9IHfrkmalOqFEpy6N/pZI1efJkOeuss/QjAH6wbNkyuw/5yCOP6BYA2SCUQGC1trZKmdX5raur0y0ASl2xzCnRLzLIrpzcJqFwgFAl7a55E8yhROQ2j8RXDaifoV9v/Yz+8MFZbtSqBCuAJJroMuNQQnpl+Tj92guif++44YtbR41+Tpk0rNZtJW7YsGFy0EEH6UcA/ECtvqH6kKy+AXiDUAKBduDAAf1/APygt7dXpkyZIr/97W/lzTff1K25Zw4lLOGrHcZI3UX9S4VOmBnvCgL3YL1LGvQVELGBQazwzy7Xr+9tCYcN1a4rMgbIKpTok9CGTmlvrJMGNQHnqAqpHh4zf0XMa1MKJbY0SLV+vV9CCXWe4VwD+A/7NeAdQgkAALIUN5RwLf9p/nqcwXpKE0rGln796tpwW8KBfYahxI6VdXL5YP31RJVJKOH6nfwSSgAAgMQIJQAAyFL8UELk/dW1MkEPtCuvbJf3dbujlEKJHc2VkckowxNVRia8fN99+wahBAAASAGhBAJr06ZNMm3aNHvCSwD+sGvXLlmyZIn8z//8j+zZY5zeMScShRKROR7GSMsW3eSSLJRIdvvGALkKJfra5RZnxY7RdbLOONEFoUSsO+64Q66//nr9CIAfrFmzxu5Drl27VrcAyAahBAKLiS4B/ymuiS5TYx6sd8ocPTifMCt6DoqkXPMy1CxOEGikG0o8Wxu+SmJOh24bgFAiFhNdAv7DRJeAtwglEFhqErxVq1bJq6++qlsAIDPehxIh6Zis20+sk026NTWd4Ukyx13krMhhkG4okdIVGIQSsZ577jl54okn9CMAfvD666/bfUg1uTKA7BFKAACQJe9Diei5KCZN74wfLlj2rGuSuludKypcy4mWjbAG93263aW3XRaOi6yWkVIo4boCY8LVht9nf690zRgTmXOCUAIAAKSAUAIAgCzlIpRQ4cK66SP018qkcmSVLFoSmVRyT0+XdDTXyYzROlyY7pofZ0uDXKJfN65skFxuT0jZLl3LmmThZFdwoCulUCI8N4aqcqmeVCfLl1nfc2WbLJ9ZLZfErshBKAEAAFJAKIHAUrdvPP7447JtW5ojCABFS01ge/TRR8sXvvAFeeWVV3Rr7uUmlFBC0VcfxK3yAXNP7GitlknG51pVPkYWLq6Xy/Xj1EIJS0+T1Di3hgyoQXJtY5ss4vaNKM8//7w8+eST+hEAP1C3bag+JLdvAN4glEBgMdEl4D8qZBw/frxMnDhRduyIGk7nVO5CCa23074a4fKRQ10BxSCZNKpKbmlsk+6d+nmxettl0eQxMkEHCZUnjJEZs9tku1o5I905JRw7u6R16liZ5FwZMXiEXDu9SdbZ34I5JWIx0SXgP0x0CXiLUAKBpT5Rveaaa1gSFACQM3PnzpXa2lr9CIAfdHV12X1IlgQFvEEoAQAAAAAACoJQAgDgG3v37rWvfnrmmWekr8+w6gQAAACKCqEEAMA3Nm7cKIcffrgcc8wxsnXrVt0KAACAYkUogcBiokvAf/bv3y9vvPGGPcnlBx98oFuBwmGiS8B/mOgS8BahBAJLXeJ9xhlnyL333qtbAADw1uTJk+Xb3/62fgTAD5YtW2b3IZcvX65bAGSDUAIAAAAAABQEoQQAwDdefvllGT58uJx++uny2muv6VYAAAAUK0IJAIBv9PT0yDe/+U0ZM2aMvP7667oVAAAAxYpQAoH19NNP25+mMqcEACBXfvnLX8q3vvUt/QiAH6g5JVQfkjklAG8QSiCw1Oob//AP/yA33nijbgEAwFtf/epX5V//9V/1IwB+oFbfUH1IVt8AvEEoAQDwjffee8+eV0LdxqGWBwUAAEBxI5QAAPjGxo0b5fDDD5djjjlGtm7dqlsBAABQrAglAAC+0dfXJy+88IK8+OKL9lUTAAAAKG6EEggs9YnqNddcY094CQBALsydO1euu+46/QiAH3R1ddl9yDVr1ugWANkglEBgqYkuy8rKpK6uTrcAAOCtYcOGyUEHHaQfAfADNdGl6kMy0SXgDUIJBNabb74pjz/+uGzbtk23ACh1r7zyipxzzjly7rnnyhtvvKFbgcJRn6h2dHToRwD8oLe31+5Dcp4BvEEoAQDwjS1btsjxxx8vp5xyCoEjAABACSCUAAAAAAAABUEogcDauXOnrFy5Ul599VXdAgCAt5577jn7Mm8A/vH666/bfUh1GweA7BFKILCY6BLwnw8//FD27t0r+/btkwMHDuhWoHCY6BLwHya6BLxFKIHA2rRpk1x77bXS2dmpWwCUOrXU7+GHHy7HHHOMbN26VbcChdPQ0CAzZszQjwD4gVoKVPUh165dq1sAZINQAgDgG7t375ZHHnlEHn30UftqCQAAABQ3QgkAAAAAAFAQRRFKfOUrX5F///d/l09/+tOUT0u9v06Zvk75o3ifg1G8z8Eo3mP/F/tyMIr3ORjF+xyMUu/vZz/7WT2K9o+iCCXUBFD/+I//aK8tT/mzvvCFL9gTAh188MHGrxei1A6tfqfy8nLj16n066ijjrL/pocddpjx65Q/6lOf+pT9Pg8ePNj49ULWscceK5/85Cft3/G4444zPodKrT760Y/KP/3TPxm/RqVeqo/zf/7P/zF+rdCl9hG1L//bv/2b8euUP+rII4+032c1347p61T6dcQRR9h/U/Vf09cLUYcccoj9Ow0ZMsT4dcof9c///M92+U1RhBInnXSSfUKEf7399tv2gfL73/++bik8NcFlRUWFLF68WLcgW/fff7/9Ps+ePVu3wI+qq6vt9/kvf/mLbikeTHTpnaFDh8pnPvMZ/QiZ+tWvfiVnnXWWflRc1LKGal8+//zzdQv86K677rLf57lz5+oWZOvhhx+2+5BqDqNi8dOf/tR+n1966SXdAj86+uij5XOf+5x+5B+EEsiLYgwl4D1CiWAo5lAC3iGU8D9CiWAglAgGQolgIJTIIUIJ/yOUCAZCiWAglAgGQgn/I5QIBkKJYCCUCAZCiRwilPA/QolgIJQIBkKJYCCU8D9CiWAglAgGQolgIJTIIUIJ/yvGUOLpp5+Wb3zjG3LPPffoFmSLUCIYmFMiGAglvPGLX/xCRo0apR8VF0KJYCCU8F5bW5vdh1RzSxQLQolgIJTIIUIJ/yvGUOLBBx+0Z0S/8cYbdQuyRSgRDMUcSoRCIVmwYIE0NTXJ7t27dSsyQSjhjeHDh8vHPvYx/ai4EEoEA6GE9xobG+0+pDrfFAtCiWAglMghQgn/4/aNYCCUCAZu3wgGQgn/I5QIBkKJYCCUCAZCiRwilPA/QolgIJQIBkKJYCCU8D9CiWAglAgGQolgIJTIIUIJ/yOUCAZCiWAo5lBix44dcs0110htba193EHmCCX8j1AiGAglgoFQIhgIJXKoubnZLviXGhwU2/usJsS7+uqr7Qkv4Q0VSrA/+5/zHjPRpb+xL3vj9ttvl+nTp+tHxUWFErzP/ue8x7zP3nn++eftPuSaNWt0S+E57zGhhL/5dV8uilACKITW1lY7Ua6rq9MtAAB4a9iwYXLQQQfpRwD8QE10qfqQxTTRJVDKCCUQWG+99ZY8+eSTsn37dt0CAIC31CepTz31lH4EwA/eeOMNuw+p/gsge4QSAAAAAACgIAglAAC+wZwSAAAApaXAoUS3NF1Qbt+TVVY2Vpq26WaUpr5e6VrSIDWTx0rFCc77atXgEVIxsVaaVvfqJxaHnTt3ymOPPSavvPKKbkEq+rZ1SUtjjVSPq5Ch5fo9LhskI0ZVSW1zp/Tu10+Er4Ra+1fcKJveqVuKk9qv//jHP8qsWbMkFArp1mALbWmTeuu4PGKw3l/Lh0rFuGqpX1lcx2S/evbZZ2XVqlX6Ua6FpG1y//tcu1o3wRdCG9qlafYUqRo1QgbZ513Xvrys23rnkU9qkljVh1T/9cz+kKxf2ST1U6ukYvig/veY9zlQupsrpVy/72Obg3WOLmAo0Sed00dEdjhCidL2bF3kJJmgBk1sku4iGbQy0WX6ehdXDnhPB9TgKmnp0S+AT/RJ+1QdNBZ5KIFo3YurEh6bR1zUJt36uciNvE50ubtdpuiwmFDCP7pmuD7oiVPlI2ulc7d+AXLO84kue1ukKvxBT/yy32eSCV/qW10rI1zvNaFEXvRJ16wx4SSovwglSpq1I/W/j+Uy9OxqqWlskfaV7Va1SMPVVVLhOtCOmNFlbQGFt3nzZnuZttWr6bmlqrd5bP/7WD5UxkyukYYl6j1ul7bmOqka6eo0Da+VrmJ4k5G1vt710jbDdbwmlCgZfc9GOjjlI6dIw+pu++qR3g3t0nBpRfg9HTOXWCKX7rzzTvnDH/6gH+XI/j7rfW2T2rMjx2FCCf/onK6Pv4NHSNXUemla1n/ubWmcImOdK6CsKr+gRbj+KT/UBLaqD/nCCy/olixta5Kx+n0cNLxKpsxukja7H91mXyHD++xvfWvqZExMKEUokQfhS1PKK6X2aj3IIZQobavrZOz0FunaqR/H6mmSSmdnK68VhjWlqXdJrdQuWR/n8kH37VhlUt1KlF/KwgFUbBV5KKFW1ZkzZ47ccccdsmvXLt0aRN3SMFq/ZyfWGD5BdV39UlYlLfGO3ShyvdI0Tr/PMUUo4R9dt1bFv91qd6fUDnfe95Okbo1uR2nZ1iJTJtVL+7Y4n+hEvc+MmXwlPEYql8rpNeFwilAixyKXpoywTph9ro4vO5jfdV7dv5PxXvvYlgap0AdTPlEvbb2tU6RiVEW4wvOHFPn7ykSXmmtfHNMY50oI13OqlhAilqZeabs0sp9WjBwavgKGUCI4+lZM6T8+WxW0gUyQdC+MfFjA/u0TrrBphNW/6nNdMUMokUPuQKJmRX8HKJtQIrSlXZqmx0wGk6BOmtWlX4lCCIcSaV0pEZLulU1SO7EiMklbwkr9U4IDBw7IBx98YP8XHtndJtXOe5HO4HW/miS1PmbyzERVLW2MofLI9WlsNqFEb5e0zK6Wsa6BU8Ka3MakXhmKzP9SIQ1bdOMAXVLr7G9T21O+rY5zb3rUeUZVXrg6tFkNWthXS8uG+vCtWmkNZDj3ZqwQfchIKJHemIljdpFyBxJXtvcfQ7MJJUr8uJ2/UMJ9aUpz5FObjEKJ0HppmJTiHzxciTpmyDnX5FsjUjzYhdY1SJV7FY9UalRDypO2MdFlDrgOpuUzUnufe5fVypiUAidXpTGAgheyDCWsjm/b9DEpTYbrrikreJczFZkYb4q0x/0zut7XE+sk6R7LuTcjeZ3oMttQgn21NIXn9SqTysWpDWQ492bH84kuk4rcklee6mCSY3YRi9zyXH5BU2Tskkko4ZPjdn5CiZ3tUuO+NEU3K+mGEn0brIGqcxAdPFZqm9tlfW/Inryre3WDTHFNtjdiVqfd3l8cRguiLyTrl9VLlX7/B01sSSE06JP1jZEZ4weNq5Wmleul134fu6WzcYpr4swRUrfaeY+tSmPm6c7OTvnmN78p9913n25BttyXkCY/2IWk/Wpnsr1yqZhULy1r+ifiC6nJFWeOdR1gq6Vlm+t9ZnfOsyxCCXX8d47L5RVSNbtFunr630c1MV/dONcnN5Nb9H7eX7zNmYosC1lWlvjKtPAEekmex7k3c9XV1TJmzBj9KMeyCSXYV0tW99wK/d6kMqDk3OuFhx9+2O5DPvLII7olR/b3WcfYJqnRk9imusoKx+xiZu2DV+oVKIfHvJ/phhI+Om7nPpSIuVcmNtlLK5RwTZY44iJrcGtaWtJ9+Tiz0xZMpKPbX4NGp76+cmSN3hFSvcQcYfQtqw5/71Q/FUCuuSbWK69J8Oms4loSuHyM1D1r3jLW3+p0tMql9lndiALIMJRwHf/Lz66TLuPbvF7qR+nvXV6b/NP6JJhTQnG9XymHEgnOwZx7S0emoUQB9lV4xHUlatnoZFeLcu4tFZHxUX+Vn1Aptc2d0pvKsvocs4uYax9UgUTsLphOKOGz43aOQ4k4l6a4pB5KuC5bivO9+rk6Y+Oa2MkKJDaUKCsbJGMm10tbT5JcbkuDjLGfH32bzwCZXN6EnIrc62i9d0mCInVFhRM8qQlv43JdksqkToWUSSjhWt0h9pOAGKl+Wp+K3t5e+e///m+54oor5M0339StQZP6eTB5KMG5t6RkFEoUZl+FF6wBztUn6ffEOp8+m7iPxbm3dMSGEqpvNfT8WmmKEyRFcMwuZu4VKJt6dKNbyuMb/x23jaHEwAFlGuXqsIYn2vrCFGswGrlcxF3r50ZCiYZ1ut3wh43snGOSXJrWLU3f178LO1kCnVJr/z0zq6Qnqd36vexdL53L3JeJJToRug6QydL+HkKJlLgObulXardUKZFJbJOdBC191rZ3on5usvtTO2rCvw8do/gGdl7SqJSOk+mHEmqbOMn+GeVJb+WJrMzDQMcb3oUSnHtLTAahBPtq6YpcWdp/NXLCd49zb2np0/3oULd0rWxxXYaf+EM7jtlFrLdFKu33ZqhMadW3TMXWuobI+Gbuet0+cG/143E7t6FEph3lAZ3eLqlzDqRXJjmQ9rXLFP19mD02kRyHEgNErpqJe2n/mrrwDlazMvEOFpm7IPM1uZ966in52te+JosWLdItPpSPUMJ1mWAq9zqGllTp75/shGltNeH7ZJnxO5HiCyVC0jJRPz/p5cTd0uBcWsgM/h5xvV9JOiGJQwnOvV6YNGmSfe95XqQdSrCvlir3hwGpzNfFuddbDz30kN2HVHNL5Ifrsv+47yHH7KKWcZ889jzuz+O2+fYN51PuTMo1IPEslLAOvP1JcHnSk2zfyhr93MwHq0HRZ3r/Uqy+VO5pi+VK3k2pXud0J7RIluT1SfuV+rmpzBgfx4MPPij/9E//JLNmzdItftRnfP9SrcTRkMUVSCS7fKxf6p/g2gdSZ44Kq8NFvyiB8CcqmVTSd9mSZijhOvEmvZIpfMtWmVQtyf5d3rVrlyxdutReXWfPnj26NWg8muiSc68nRowYIZ/4xCf0oxxLN5Qo4L6KzKV1daKNc6/X5s2bZ/chFy5cqFvywHWVcMVcw7vOMbu4eRVK+PS4nfuJLpOIBBfxP5WNLG1Wk3yw6txfkzQ5Qt4lXEfbtV7+1UkGPa5JncY08i4XTE9LZGbnlAIJS/jSNWsbWJjkvXPNL5LsyhnkWnqhRPjWPXVcN90z6dI9d0z/c5NOjpoaJrrsl9p50/W+GgJezr0lKM1QopD7KjKjAglnBbLUAgkL515/cE9OaTgXc8z2gRQCB78et0sglEgj3XUdSKtbyXaLjnUi7X+vDTtaGqlfZAerlraduhH51dMUDiRSXZ7K5toGEneY3RP4ZH41DLySXiiR8NN3N1fAOILLRz0VanVWKEq0RKArDB5wWSfn3pKUZijBvlpaQh2RQCK1JdY1zr3+4Nq/B56LOWb7QgrjIb8et0sglHB1mhLuZN3hnTHl5Bh51TXTmSHasMTUs84lZ0lCifCla0lW50DO9K2pkzF6nyy/oEHWpxpIWCLpbuKOUV9HjZ5fJMkM4ciTdEKJXmm5QD834QnTNWt8qlfaIHXhTmeCK8qs51To5ww87nLuLUlphRLsq6Wku7U6fLXpiCvb07qtgnOvP0TmBTEth88x2xeShhL+PW6X1pUSasId0x91v7WDTdSz0sZbYgU5ZO0gl06RhtW91i5gptJ952RqTN/d6e9Fbebv4/503oMD6YYNG6Smpsae8BKpCa2oiXSKrPcp7ffA9WlNxa3rdWM09ydBSWcTR55keqVEhdRv0I1RQtI5o0IHkXR+c8P1np1YY+iQuD4RLauSlgFXnXHu9cqcOXNk2rRp+lGOpRVKsK+Whj5ZP9dZZaNcxszqSv+8yLnXc88//7zdh+zq8uYT6M6ZY6Vu2XoJxZuzzT2Hl/FKYY7ZvpDWlRL+Om6XxJwS62c7s81ag9GRavCrl1Hp6ZK25loZ69zXPtjqWLGDFYDrQDh4hFRNrZOGJe3SvtKqJQ0y5fyh4asg4u8c66V+uPOccqm4tEE69TKy3WvapGn6WBmkv0dalywmoCbCU9+vrq5OtyAx14otZ9dKi3p/k1TXtpj3OtQm1c5JtWyQjJ3ZJut71fvcK+tXt0j9JOcgam0DM+gUFY/0QonQsurIPj+4v6PVq47ZaongJfVS5SwRXF7h+cly8+bNMmjQIBk6dKi8+uqrujWY+p51TYbnOnf2bmiXhosi59V4l3Vy7vXGsGHD5KCDDtKPcizNUKKQ+ypS5HpPh05qMJ5rY2t97ICVc6/nGhsb7b/nggULdEt2wgPN8qEyZnKN1De39b+fy5qkbvKYcB840ZXCHLN9IIVQwq/H7ZIIJWS3NRgKD1hNpQaxTbKe26IKpFdaJjqfuCUoa+eoWRH/TXLPJm0s6/VTmtendcliIm+99ZZ0dHTI9u3bdQsSS38ZWdMB1b2uurHUAXZl/AsPkR+RY3PiMg98uiNLAMepQePqpD0Hb/O2bduksrJSJkyYIDt27NCtwdW9uMrVmR1YCUNezr2eWLt2rXR2Jg/zMuP+dDRRxbvMt3D7KlLkGqSkWqbjMudeb73xxht2H9Kr80zXTH0FQ8IaJFWN6+OHRhyzS18KoYRfj9ulEUoou7ulbXa1jB0e2WkHDa+QqulN0hn7aSwKoE9CG9qlafYUqRpVIUPDiXy5DB1ZJVMa26Q7hYNgX0+b1E8eKyOcNNc6AI8YVSW1zZ3Sy9tcYN6EEkro2SapnejaTsqHSsW4aqlPdOki8iq7UEIJSVdzrVSNjFwpVX5ChYydXC9tG+gR5VNoi+G4qva3VAYgnHuLXLahhMK+WtQ8CiUUzr3Fra+nS1oaa6R6XIXreK2OuWOleqZ1zE1lkMkxu7SlFEoo/jtuFzyUAAAAAAAAwUQogcDauXOnPProo7J161bdAqDU7d27V1avXi3/+7//K319fCqEwlPbYnt7u34EwA9ee+01uw/5+uuv6xYA2SCUQGAx0SXgPxs3bpTDDz9cjjnmGAJHFIW8TnQJIC+8nugSCDpCCQSWmqW/trZWnnnmGd0CoNTt379fent77UnIPvjgA90KFI4avPzxj3/UjwD4gZrAVvUhX3jhBd0CIBuEEgAAAAAAoCAIJQAAAAAAQEEQSiCwDhw4IO+//758+OGHugVAqevu7pYRI0bIGWecYU9EBhSauqVInWsA+IfqO9KHBLxDKIHAYqJLwH96enqkoqJCRo8ezazoKApMdAn4DxNdAt4ilEBgqWUDR40aJffdd59uAQDAW7/+9a/lO9/5jn4EwA+WL19u9yFXrFihWwBkg1ACAAAAAAAUBKEEAMA33nvvPXteia1bt9r38gMAAKC4EUoAAHxj48aNcvjhh8sxxxxjBxMAAAAoboQSCKynnnpKTjvtNFm0aJFuAVDq+vr6ZO3atfLCCy/YV00Ahfazn/3MnnwVgH889NBDdh9y2bJlugVANgglEFjqhPLP//zPctNNN+kWAAC8deqpp8rBBx+sHwHwg/nz59t9yIULF+oWANkglAAAAAAAAAVBKAEA8I1XX31Vvv/978t5550nb7zxhm4FAABAsSKUAAD4xpYtW+SLX/yinHzyybJt2zbdCgAAgGJFKIHA2rBhg/zud7+Tjo4O3QIAgLduvfVWueaaa/QjAH7w3HPP2X3Irq4u3QIgG4QSCKzW1lYpKyuTuro63QIAgLeGDRsmBx10kH4EwA8aGxvtPuSCBQt0C4BsEEogsN5++215+umn5bXXXtMtAErdgQMHZN++ffbSoOr/gUJTy9OuXr1aPwLgBzt27LD7kH/72990C4BsEEoAAHxj48aNcvjhh8sxxxwjW7du1a0AAAAoVoQSAADf2L17tyxfvlweffRR+4oJAAAAFDdCCQSWuuTukUce4dNUAEDOPPPMM/LYY4/pRwD8YPv27XYfkluAAW8QSiCwmOgSAJBrTHQJ+A8TXQLeIpRAYG3evFlmzJhhf4oFwB9ef/11mTx5svzmN7+RnTt36lagcObNmyczZ87UjwD4gZrAVvUhX3zxRd0CIBuEEgAA32CiSwAAgNJCKAEAAAAAAAqCUAKB9eGHH8p7770nH3zwgW4BAMBb77//vn2uAeAfTh9S/RdA9gglEFhMdAkAyDUmugT8h4kuAW8RSiCwVq9eLd/61rfk/vvv1y0ASh1zSqDYXHTRRXL22WfrRwD8YPny5XYf8tFHH9UtALJBKAEA8I1QKCTz58+Xu+++W959913dCgAAgGJFKAEAAAAAAAqCUAIAAAAAABQEoQQCq6OjQ0aMGCHNzc26BUCp+9vf/ibTpk2T66+/Xt5++23dChROVVWVnHHGGfoRAD948MEH7T7ksmXLdAuAbBBKILAeeugh+ehHPyo333yzbgFQ6pjoEsXmtNNOk0MOOUQ/AuAHau4i1Ye86667dAuAbBBKAAAAAACAgiCUAAAAAAAABUEoAQAAAAAACoJQAgAAAAAAFAShBAAAAAAAKAhCCQAAAAAAUBCEEgAAAAAAoCAIJQAAAAAAQEEQSgAAAAAAgIIglAAAAAAAAAVBKAEAAAAAAAqCUAIAAAAAABQEoQQAAAAAACgIQgnkXGhDuzTNniJVo0bIoLIyKVNVPlQqxlVL/bJuCenneWLnemlrVD+rQoaW659VVi5DR9ZLl37KAH290q5eM3KolIefXyW1zZ3Su18/J66QdC+rl+pxrp+n/211S7oklPT1QPHoXd0ktRNd2/LgEVIxsVZa1nm4l1r7W2dznb3PjBisf45Vg4ZPkbZe/RxbrzSNi3w9tRorTdv0yxPqlqYLyhO/ZluTjB3w/eNVrXTqlyE7bIMubIMIgv0hWb+ySeqnVknF8EHhbbr8hAoZO7le2rZ42Uvss48xdZPHRv0sdZyZ0hq180cJrWmxXxM+XljPT/l36+2UpunR/7ZBwyukamqDtG/r008CQCiBnOqa4XS64lf5SKsztVu/IGMh6Zo9NhJ6DChzh61vTZ2MCYcXhhpcJS09+smxQp1SOzLxv698pNXJjfd6oGj0SeeMCh3Kmapcxszqsp6Vne4lU6Qi7v4WOzDL1YDQ+rdOH5H8NQwI84xtcAC2Qfhel9Qm6oPZVS4VMzqz3velp0WmJOizjW02hRLd0naRe1+NrcS/W3dzVYJ+qapBMnZ29sc1wA8IJZBTndP1gXfwCKmaWi9Ny9qlfWW7tDROkbGuT6jKL2ixun+ZijlpDB4j1Vc3SIv1c/p/Vo1Uj64b2GHraZLK8KdxY+0rG7pDIQn1rpe2ma6AY7gpNOmWhtH66+UVUrNkvYScs0pfSLqaq2VE+PV18a/SAIpA7+LK8GBw0Lg6advQKyFrX+hWnw6Ncz7dKZcpKzLtOvVJ16wxkQFn+VCpdB0P2prrZcr5VdISNTDrk95n+7+erNT+2v87JhsQxvweiV7jGhBWXN1i/LmRsvZ//TJkhm1Qf9mNbRC+1ym1ehsfNLxKpsxukjZ7e26RhqnuD5rKpXJx5r3E6A+gymXo+VOkvrmtf99Zpq7SqJSqAd/fHR6WS8WlDdLuHJdWN7gCDut3a+7Wr4noWzElckybWC+dPZFjV19Pu9SeHXl9dSt7L0AogZzqurVK6lfGOZHstk5Gw/sP2GVlJ0ndGt2epu65Y8IH9jEz2lO45UJxfQJWPkXaDOeD0MoaOcn+vmVy0szoWEGdbJyfGa+T3NvsdFLLsuhIAznW1y41TmdtXJMM7Fq5ArgTMwvY3J2zERc1yXpP+1+u3290g+H3j+hu1gPf8kqpvTrJINI1IDR/ggbPsA2yDSKguqR+Ur20x9m8+1bXRj7gyXDfl93tMsU5vgyvlqYUbwVTP9vpA465db1uddlv7ffhK6mqpGWnbrdZXxulvxbvmNBn9YFP1M8Zlfi4AQQBoQQKKjK4z7DT5eq0jZiexuV9WxqkQr9uTGO8U0FIWibqE0Z5jbS7vnnktpQpUe1RQm1SrX/GSbO4VgLFyR2w1aw0b8yR52QQHro6XuUXmAac2YkMNuP//kqkcztCalf3uUJDBoSFxjbINgiY9Un71P59IO5+kpC62uGk/teXV0pTyrfTun5uTP8vypq6SHDh7kv2tkilbq+YG/+I0zVL/25l1cYPx4AgIZRAYW2oD6fgmXS6wgf0RCcNA3WpcP+JIHEHN9RarZ9XLrXP6kZL+LaURCdJVygxdmGq3eCQdK/sn+jNPQFb/Mr8ChNASSlgS7GDZRLZh8ZIwxbd6JkuqXOuthrXFPcWMPdgsGZFf8+vYAPC/b3StSRmctyE5f/OKttgnrfB3i5pmV0tY8OTOyepyW3cGoKCWX+rcwtFBqGEqx8W/wMoE9dcFwm3f2v/d652cN+G7Np3y6fHn+0lEkrUpD4nDPsvfIpQAoVlddScA2f69wtGTgaJDvomSTuDDtcVFe5OYaSTm+AKjXCCnlpwEFrXIFUnOJ3zFItL/pCVXmm5IJVtKXLfb3odnJC0TdavSzBgy5TaD/s7ZSfZnzwbheeOib7vtxADwt5ltTImpbDRVVPbU78CrCSxDeZtG9zfK23TxySZeG9gcfshCinyIVCltKS5G0T6aukGGq7jTcL+pftKDvdks66wIu4VGq6rcVM5prH/wucIJVBQ3XMr9IGzIv1PsFwdturWXgltaJN695JNZYNkRJxlRyOdwSQ/1/Uzoj+d65aG70cCBDUxW/TSTpHl3sqTDir6ZH1jZIbmQeNqpWnleulVk26GuqWz0T1b/AipW63adWW9agmCzdVxSjhgc83BktbALtKxU7cw9W3TS6OFw7d0lt+N4b4fd2KLuUO3s11q9KfYseFhOgPC6FK/81ipnt0m3SmPjEPSfrWzskS5VEyql5Y11nFJ7cOxE+uWVUvLNtc+7vv+JNtgXrZB9Xs4E/OVV0jV7Bbp6unfxnqtc2dkMlGrJrfo809/+X4TRBFzzc2QwYcw4UBDzUdhLwVc61r+3eqfnWDtC9ObpHPAAcUVSiTpw0VCk+grvfpW1kSO69Y+N6U5epn4yNVTKVzBxf6LACCUQOG4Jx9KMjmYkesqi0GDXQdkQw2a2CDrXQN495UOVUsS9OrcncLYtHz3emmY6P65akZnFSi0S5NeDST255qEJz4rGyHVS8x/hb5lkd83mxmogWiujleqA8J0lh507T/l1j7qdARNVT6yRtqjJgpLLDKgG2G+Esk1ka4aDMbu5ZkPCF1ldQ5rO5KNCl0zuJePkbpnzc9ff6sT0EbfKuZ/bIM53wZdv0f52dbgzPj09VLvDP7Ka4VZkFAMIvO1pHv7heKe0HyQDAp/uGMoaz9ybqvq5wpLk0ywmeh23tCKmugliNXqcLNbpHOZXg1kwM81YP9FQBBKoECsjvrVzr10I6xOePpZbmReCFWDZMzkemlZ7Vxh0CvrVzZJTXjJJetg7r48bmeLVDmvNS75qfU0xA8llN628KdgA+r8BulO9snblgYZYz/fvKRUmKtz6um9xQg414DQtH2HZTggfLbWNQjUod0yvfSuVWq5x/qJruV8U10+1xVoRu3XYa4rleJMbJh8QGgdI86ulrrmNunUy8DZx5XVbdIw1X0JbeJPuSKd6v7JDeNyhay1q3VbILAN5nYbVJeX6/NgonOdJTK4SuPvC+SKNRivcYIBa9vtSrub6JoXwqryEyql1tqXnCsMQj1qboaq8LxmseFi10ynj5q4f9Z+pfN6034ckq65lTI0/DPcZZ0TViYJJNh/ESCEEogS6SRlUGlcUhu5OiDBnAxJuH/XmpW6cQCrYxjuyLrnduiTrhmRjmj5yCqpX+J0VFWnT62RbZ1IXCe02MkquxfrWy7KK6VhXaj/suCZ1dH3jA8eK/VxPhmN6mQnu1Kkh1ACWiqfnsat2E5TjgeEroH22LifcrkDSnUrVrJOmrXvhZcBNg/GwoHlF6ZIm9MBjan1cyMDQrX/2u1JrmpyCy2bEp51Pe79wO5VH5LdxtVRo38fQgkztsFYqWyD6hLx/ufEX77a0Xm1/l4MalBwkVAvaaAbl+vY8v34faw+69hr3I+2tei5YFQNkrHTm6TdCQd7uqStuU6qR7uvlo2ZrFJd4aBvuRhxZZv07rb2+WUNMuV89wSV5VJxaUvcD7DYfxEkhBKIko9QInIfnbq9wToY6/Z0Jf2UyWH9POcEEDsvRJu+zSJeDRpZET5ZuTuqfc86/4aTpKYj5kSxPyRdzdHzQBivBAlPhJl4GTklq+Xw4C+5CiVycem8a0CYcKDt/jclG7zvbJNq5xPqOM/N+DiWcFAcy/03Mc+cHlpSpb+e/J7hyPw6QVsajm0wqjzdBl0T6SW9RdJ17368kA3IC9ctb2WDpGpxpr3ETALPmHkhrH6afZuF831iq7xCKpyrZaNu81AT+OpQxTquxf4L+ra1R80DUe5euSOM/RfBQiiBaH2RT3HSr4TdOJs7kMh2vfhIpy/JPdiupeRMJ6be1f1LcIaviigfKhUTa6VptXWKCF/66w4DXCewBCcK+2Tm/FxDZ7tzuj5hlSfrYPdJ+5X6uUnubUQQ9Bn2vdQrei/N8SSDrgFh4rlQXJfZJvn+kUtq4w/08zMgVPuw81pTMJrO38zqUI7Wz403YaJvsQ1GlZfbYDq3/YVvJUwyzxKQU+5AIsltrUm5QgnjoD8isiyxYT8KdUubWoJzeCREGDTcmWjWtXqQOwywjjv9fcdEHzqFpGWy83OtfmxsaMr+i4AhlEDeqEDCuXog20DC5ups1nToNqNOqdHPS7fDF+5YuoMD17rXydL38MlqwPr7rg7w1Ul+J9e9y+lP9AQk4tpGv59on3R17tL5FMbdqYq5/SlatzR9X3//RANC1/c7ydr3kseg8UUGjUmutEog4YDQFYYm/rdbXHPLJLtqyn/YBnO1DUbmXbK+ZlySMCJ8O0p5Tcy5CsiXPumcEVmlKLtAQnEFmQmPLdb2vzDTfdE6Lun+mTs46JrlBJdJvp9rzpvoK3nZfxE8hBLIi1BHJJDI5paNKK7OYezBPIq7EzkrnesMIp9ell/pukTX9f2ShRyRDmPM+trujnKSBDxysqmWtjRmhgdSEek8xQZnLq4Bdnr7UKTDlviS+FQGnK4rhtLuOA6U/YAwJG2T9e9sunTeFZomniPCPZFZMK+EYhvMzTYYOf8kuRrPFXyPSOtvC3gl5AoksrllI1r4itRExxZLZF9J7/Y5texn/+8cfdVUwsDazdUXLJ8Rve+x/yJoCCWQc92t1eFbNkZYg/s0jvdJuC55TnBbQ2Q+hvQ+hXQv1Rk9j4PryotULzeOXaLJlY4nDCXCE1x68akBMFCkUxV//3DvQ+lNwugexFVJS7xQbUuDVNjPSXA1kOvy1JNmZt/xynpA6F7Bx3DLhXt1oER/s8gka5lO5lb62AZzsQ26r9RLNKhxTfKZZHZ/IDfc83uNSL5EZhoix5ZEtzW45mNIZ3l691KdMUFm5MqLJMer8G0esbeXsf8ieAglkEN9sn6uM7AvlzGzuqyW9EQ6bVYZrkqITCQX59Jc10kjnU8hu5e4gpQBl+i6O7mGiS419/wZAzqw7qstLmoz/116mqRKr+Thye0ugIlrhQgVsg3czpKEf64rAowhXXhCV2s/uNp0ubtrlvW4VwOpT4STPSc9iQeEnVI3rlZa1sTvxEZmho8zK7rr71Jx63rdGM19BVmmqxD5AttgTrbByCetFVK/QTdGcX86HdxQDAW0e700hPe9MVK3Jt1t0HWFk3E/cs1Zc2KNcdAe+QCqPKWVd2wh6/s6S86XVw68vcIVYJqPaYp7Hx4YmLL/ImgIJZA7roH30EkN0r6yPWmtjzkoJwslYjtmFZc2SKe9/FqvrF/ZIFP0ckzmA3avtEyqlCmzm6TN/vlt0jR7ilSe4LwmQRjQ0+RaKmqEVDd3SXiezz617FNtZMZmY3q9XuqdsCTq91br5lu/x/Sx4TXoPbvdBYhDfarvfFozaFydtOllz7rXtEit0/GK12FLNiC0hoCRicsGfv/IDOQJrgayBpXhkNCjy1OTDQjDHd3BI6Rqar00Les/RrU0xh4j4kygpuaeCR8jBsnYmW2yvlcfm1a3SP2kyL3TFTMCHEhobIO6MSz7bTC0rDr8N1XLU9ctWy+91r851LteOpfUS5VzfiyvYECDgojsA0Olam6kLxi/1kddkZA8lLD2fuv44Oy7UftBT5e0zIz0tYz9vdV1MmZyjTQs0T9/SYPUTB4Tfk38MCD6mFN+dq2090Se19fTKQ3hq0PMxx32XwQNoQRyx301QIoVe5lb8lDC4loL2ljWAdt8OaDr9ooBZQ0Urm6T3jhrRyt9GxrCVzLEq/KRNdIe5xO1qBOlqazfe0pz7AkYyAWrAxX+xMVUCQbOSQeESre0THSv5x5bg6SqcX2cgblrWbTyKdLu0eWpiQeEroloE9SgiVYnNsExIvIJXJxSHc2V5r9Y8LANRvNiG3SH9uZSAU07myAKJKqPl1LF3sqQPJRQuhdXuYKEgTVoYoOsN+3X7mNLbA2ukoZ1iXpoIWm/OtExTVWi4w77L4KFUAK5k69QwhaS9UvqpHrciEjqfUKFVE1vka64l9n2StulkeerTu/QkWOlemaTdG4znyIG2N8rnc3RP7d/SdEp0rCsO2mg0NfTJvWTx8qIcLgxSEaMqpLa5k7pTfFXALwSb3nclkQdr5QGhP161dVL7u8/eISMnVwv7Qn2NxXeOZfej0k0oW2aEg8ILfutY8rKJqmz9s8K11Jw/b9zXf+SwSkIPWv4m46rlvpl6yWUINAIKrZBF0+2wZB0NddK1cih4cGROjeqf3PbBiJvFFa+Qglbb7s0TK2SivBVRlZ/Sx2LV/bGCQUsz9a7nm+Vte/19+9SP36HtrTF/Nyy/iVFVV8z6S7M/ovgIJQAAAAAAAAFQSgBAAAAAAAKglACAAAAAAAUBKEEAAAAAAAoCEIJAAAAAABQEIQSAAAAAACgIAglAAAAAABAQRBKAAAAAACAgiCUAAAAAAAABUEoAQAAAAAACoJQAgAAAAAAFAShBAAAAAAAKAhCCQTW3//+d+nt7ZU9e/boFgB+8Le//U127typHwGF9eabb8obb7yhHwHwg3379tl9SPVfANkjlEBgtba2SllZmdTV1ekWAH5wxBFHyJAhQ/QjoLCGDRsmBx10kH4EwA8aGxvtPuSCBQt0C4BsEEogsJ5//nn50Y9+JA899JBuAeAHkydPlksuuUQ/AgqrpqZGfvKTn+hHAPxg1apVdh/yiSee0C0AskEoAQAAAAAACoJQAgAAAAAAFAShBAAAAAAAKAhCCQTWypUr5bjjjrMnKwLgH9/4xjfk29/+tn4EFNa4cePkK1/5in4EwA9aWlrsPuTSpUt1C4BsEEogsFasWCHl5eUyZ84c3QLAD0466SQ59dRT9SOgsEaPHi2DBg3SjwD4waJFi+w+5OLFi3ULgGwQSgAAAAAAgIIglAAAAAAAAAVBKAEAAAAAAAqCUAKB9cILL8gvfvELefTRR3ULAD+44oor5Oqrr9aPgML6wx/+IL/+9a/1IwB+8PTTT9t9yNWrV+sWANkglEBgtba2SllZmdTV1ekWAH5wxBFHyJAhQ/QjoLCGDRsmBx10kH4EwA/Uym2qD7lgwQLdAiAbhBIIrN27d8vGjRvl7bff1i0A/OCll16S7u5u/QgorK1bt8rmzZv1IwB+8M4779h9yF27dukWANkglAAAIID6QiHZ+vjjsvGBB+TxadPCdc8558j8M86IqgcmTIh6jnqdKvU9AAAAskEoAQCAz/WuWSNr58+X5ZdcYocM08rKPKsbDjnE/p7qe6ufoX4WAABAqgglEFjbt2+X5uZm2bBhg24B4AcPPPCAPPjgg/pRMIV6euSZm2+2r3pQoUFskFA7+Csy7avfk2nnX9NfM1ZF6p6QTLP+fFF1Z0/0c5zXnXB6//eK+f7qZ6qfrX4H9bsE2SOPPCL33nuvfgTAD15++WW7D8mtgoA3CCUQWEx0CfhTUCe6VFcoqKsVbj766KiA4LrPfE6mffNCmXbJPJl2y5qBgYNXpcKKn91k/yz7Z7p+h9tOPNH+3YJ4FQUTXQL+w0SXgLcIJRBYavKx22+/3V4aFIB/3H333XLPPffoR/6mrkKIDSKmf/zf+q+CUCGEusLBFCDko9TPVr+D+l3076ZK/a5BuoJiyZIl0tDQoB8B8AM1yaXqQ27atEm3AMgGoQQAACVGzd3gnhvCDiLU1RA1S80BQTGU+t2s39H+XfXvrW7xUBNtAgCA4CKUAACgBKiVLtTKF1FXRXzp9P6rEUwhQLGWmrPC+p3t313/O5yrJ1jNAwCA4CGUQGD19fXJ66+/Lnv27NEtAPzgjTfekB07duhHpc+5RcOZsNK+0uC7Fxf21gyvSv0bXFdPqH+jCl78FE7s3LlTent79SMAfrB37167D7lv3z7dAiAbhBIILCa6BPzJLxNdqjDigQkTwlcTTFeTR6qJJE2rY5R6qX/T+dfItR/zXzjBRJeA/zDRJeAtQgkE1vPPPy/nn3++tLW16RYAflBdXS2XXnqpflR6nNs0wldGqDCi1G7RyKbUrR169Q4/hBNXXXWVXHjhhfoRAD94/PHH7T7kk08+qVsAZINQAgCAIqHmVQhsGBFb1r/dfeWEmtwTAAD4D6EEAAAF1rNqVXgCS3t+hfOvMQ/Ug1b6tg5nzonbTjzR/lsBAAD/IJQAAKBA1LwRallMNeC2S01g6cc5I7ItPSGm83dSc22wUgcAAP5AKIHAWrlypRx77LFy55136hYAfvD1r39dRo0apR8Vr6hbNdTymLesMQ/IqUjNWCXXHvMV+29WKrd0/PCHP5Qvf/nL+hEAP7j//vvtPuTSpUt1C4BsEEogsFasWCGHH3643HbbbboFgB+cfPLJctppp+lHxUddHTH/jDP6wwh1W0KQ543ItH52U/iWDvW3VH/TYvWd73xHjjnmGP0IgB8sWrTI7kMuXrxYtwDIBqEEAAB54r46YtpXv8etGtmUuqXjhNPDV02ovy0AACg9hBIAAOSY++oIe0WJmqXmgTaVfll/S2eVDjU/B3NNAABQWgglAADIoQ1Ll0bmjhjB1RE5qZirJtTfHAAAlAZCCQTWCy+8IJMmTbLnlgDgH5dffrlcddVV+lHhqE/s1SoRdhjB3BH5qZ/dZP+9VS2/5BL9ThTWjBkzpLq6Wj8C4AdPPfWU3Yfs7OzULQCyQSiBwGptbZUyq+NaV1enWwD4wRFHHCFDhgzRjwqjd80aue3EE+3Bsb1ahPok3zSIpryvW9aEV+hQ70GhJ8EcNmyYHHTQQfoRAD9obGy0+5ALFizQLQCyQSiBwNqzZ49s3rxZ3n77bd0CwA+6u7ulp4AD0TXz5kUms/zuxeaBM5XbUrfIfPNC+z0o9O0cr776qmzZskU/AuAHu3btsvuQ6r8AskcoAQCAR9QtA2ogbN+uwWSWha9L5oWXDn182jT9LgEAgGJCKAEAQJbU/BHcrlGkpW7n+L+fs98bVucAAKD4EEogsLZv3y5NTU2yfv163QLAD5YsWWLPGZMvav6I8O0a6pYBVtcovlLvyVe/Z79HKjxS71m+LF++XO655x79CIAfvPTSS3YfUt0uCCB7hBIILCa6BPwpnxNdRs0foVZ+MA2IqeIpNceH9V6p96xn1Sr9LuYWE10C/sNEl4C3CCUQWK+88orMnTtXXnzxRd0CwA/Up1f33nuvfpQ7ao4CNcBl/ogSq0vmybUf659nQoVKufbAAw/YAxgA/rFx40a7D6kmuwSQPUIJAADS9MCECfag9rrPfM6es8A4+KWKt9Q8EzqYUO8lAAAoHEIJAABSNGBCS+aPKN26s6f/PdTBBBNgAgBQGIQSCKy+vj57ssvdu3frFgB+0NvbK2+88YZ+5B13IHHdKCa09EWp9/CE0+33VL23uQgm/va3v8lrr72mHwHwg71799p9SPVfANkjlEBgMdEl4E+5mOhywAobpgEuVbql3lMdTIR6evS77g0mugT8h4kuAW8RSiCwurq6ZPz48dLW1qZbAPjBr3/9a5kyZYp+lL2oQOKSeeZBLVX65VqZw8slQ3//+9/LxIkT9SMAfvD444/bfcgnn3xStwDIBqEEAABxRC35SSDh/1LvcQ6CCQAAEB+hBAAABiqQUAPU69SSnzNWmQexlP9KBxN/OOSQvCwZCgBA0BFKAAAQwwkk7GUjWfIzeDVjVXjJUIIJAAByi1ACgdXe3i6DBw+WO++8U7cA8IPTTjtNzjzzTP0ofQQSlF3We+9FMPGDH/xATjjhBP0IgB/cd999dh9yyZIlugVANgglEFiPPvqoPUv/7bffrlsA+MEpp5wiX/va1/Sj9BBIUFHlQTDxX//1X/bgBYB/3HPPPXYfUoUTALJHKAEAgOWBCRPswWft4K8QSFCR8uiKCQAAYEYoAQAIvKhA4p6QeXBKBbdUSEUwAQBAThBKAAACjUCCSqkIJgAAyAlCCQTW2rVr5Wc/+5k88sgjugWAH1x22WXyu9/9Tj9KjECCSqsyCCauv/56mTx5sn4EwA+eeuopuw/59NNP6xYA2SCUQGC1trZKmdWxrKur0y0A/EBNPjZkyBD9KD4CCSqjSjOYGDZsmBx00EH6EQA/aGxstPuQCxYs0C0AskEogcDas2ePbNmyRUKhkG4B4Ac9PT2ydetW/chs9U03EUhQmVcawcS2bdvkpZde0o8A+MGuXbvsPuS7776rWwBkg1ACABAoahBJIEFlXbewKgcAAF4glAAABIYTSNiDSQIJKtsimAAAIGuEEggsdUntXXfdJevXr9ctAPygpaVFHnjgAf0oIiqQUJffmwaZFJVuJQkmHn74YVm0aJF+BMAP1C1Zqg/58ssv6xYA2SCUQGAx0SXgT6aJLgkkqJyWK5joWbVKb3X9mOgS8B8mugS8RSiBwHrllVekoaFBXnzxRd0CwA+am5tl8eLF+hGBBJWn0sHEHw45RHrXrNFbn8j//M//yDxu7QB8ZdOmTXYfcvPmzboFQDYIJQAAvqUGhzdYg8TpHyeQoPJQNUvtAExtc+5gAgAAxEcoAQDwJSeQUIPEaTNWmQeRFOV1XdJ/ZQ7BBAAAqSGUQGDt27fPnuySNaYBf3n99dflhUcfjQQSapBoGjxSVK5KBxM3H320vLJ5s2zfvl1vnQD8YM+ePXYfUv0XQPYIJRBYTHQJ+NPhhx0mn/7nfyaQoApbP7vJ3gaP/td/ZaJLwGeY6BLwFqEEAqurq0t+/OMfy7Jly3QLgFLXFwrJyE9/Wk5VgYQaFJoGixSVr/rmhXKGtS2eduih9rYJwB+eeOIJuw/Z0dGhWwBkg1ACAOALatB324kn9l8hYQ0GjYNEisp3qW3R2iYfmDBBb6kAAMCNUAIA4Atq0KcGf7PHnGceHFJUgarupFMJJgAAiINQAgBQ8pxAYuaQ42Vp5w7jwJCiClX3r9hsb5tqG111zTV6qwUAAAqhBAKrvb1dPv/5z0tDQ4NuAVCKVt/UP6GgGvSpwd8XvvSfMugr3zQODikq3zV0xFg5atBxUcHEmnnz9NYLoBTdd999dh+ypaVFtwDIBqEEAuvRRx+Vo446Sm6//XbdAqDUqMGdGuTdcNhR9qBPXSUxeOiJ8rnjv2EcIFJUvuu4YWfL4UcdY2+bzQva5bqPH2xvsxuWLtVbMYBSc++999p9SBVOAMgeoQQAoCT1rlljD+5qrUGeGuypQZ9TpsEhRRWq3Num2lbVNnvDIYfY2zAAAEFHKAEAKDlqMKcGdSqUiA0kVJkGhhRVqIrdPu+suaX/Ch9rGw719OitGgCAYCKUAACUFLX0581HH20P6tTgLnbAp8o0MKSoQpVpG3WCCbWMrdqmAQAIKkIJBNbatWvlpz/9qTzyyCO6BUCxU4M3NYhTg7m5F19nHOx97/zJ8vVxVxoHhxSV76r4ca18+/sXGrfV+nMnhYMJAKWjo6PD7kM+/fTTugVANgglEFitra1SZnUG6+rqdAuAYnfPOefYg7jZY84zDvJUferTh8mhhw8xDhApKt91xLHD5CMfPci4rapS27LaptWytgBKQ2Njo92HXLBggW4BkA1CCQTW3r175eWXX5Z33nlHtwAoZmrQpgZvN558qnFw59TcJc/LpY2vGgeIFJXv+u/52+W2+581bquq1Koxfxzcv1Toqmuu0Vs7gGL27rvv2n1I9V8A2SOUAAAUPWfpz5lDjg8v/ZmoTINDiipUmbZRd6ltesZhR9rbuNrWAQAIEkIJAEBR27B0qT1YU8sophJIqDINDCmqUGXaRmPLWSpUbessFQoACBJCCQTWtm3bZOHChfLXv/5VtwAoNs7Sn7WfONi49KepLr/+TvlRzVLj4JCi8l3jr2mTS6fNMW6rsXVXfX8Ap7Z5ggmgeG3ZssXuQ7700ku6BUA2CCUQWEx0CRQ3tdKGGpypQVq8pT9NxUSXVDFVsokuY4ulQoHix0SXgLcIJRBYr776qtx5552ybt063QKgWKSy9Ge8mjL9djn3t/cZB4gUle86/+pW+c3VfzZuq/HKWSp0/hln6D0CQDHZvHmz3YdUV0wAyB6hBACg6DgrbSRa+jNRmQaHFFWoMm2jyeqmr59l7wMsFQoA8DtCCQBAUVHLIqrBmFppwzRYS6VMA0OKKlSZttFkpSZ1VfuA2hdYkQMA4GeEEgisffv2ySuvvMIa00ARcZb+/NNxJ6S80kZs3fk/L8iUBa8ZB4cUle+6bGGv3PHAGuO2mqzUPnCdXpGjZ9UqvZcAKLQ9e/bYfUj1XwDZI5RAYDHRJVBcwittWIOwVFfaMBUTXVLFVOlOdBlbzlKhrMgBFA8mugS8RSiBwFpjde5+8pOfyMMPP6xbABSKe6UNtSyiaXCWav3XuZPktLFTjQNEisp3nVE5Tc48e7xxW021WJEDKC5PPPGE3Yfs6OjQLQCyQSgBACgo90ob6Sz9mahMg0OKKlSZttF0a07VZfY+woocAAC/IZQAABRUtittmMo0MKSoQpVpG82k1D6i9hVW5AAA+AmhBACgYJyVNm48+VTjICzTMg0MKapQZdpGMylW5AAA+BGhBALrsccek0GDBskdd9yhWwDkk7PSxg2HHZXxShumOu6E/5BBXx5pHBxSVL7rC8PPkSOPPta4rWZS97Y8x4ocQIEtXrzY7kO2tLToFgDZIJRAYD366KPy2c9+VubOnatbAOSLVyttmGrIF0+So0843ThApKh813Ff/a4c8dnPG7fVTEvtM3agZ+1DoZ4evVcByJd7773X7kPef//9ugVANgglAAB5pSa2vPnoo+1B1fwb5hsHXdmWaXBIUYUq0zaabbEiBwDALwglAAB55ay0oVYTMA22vCjTwJCiClWmbdSLqj93kr0v3XPOOXrvAgCg9BBKAADyJhcrbZjKNDCkqEKVaRv1qupOOtXep9SksQAAlCJCCQTWmjVrZOLEibJ8+XLdAiCXnIkt1eoBXk5sGVvf/dEv5Ws//K1xcEhR+a6RP54u3zrnJ8Zt1YtiRQ4g/zo6Ouw+5FNPPaVbAGSDUAKB1draKmVWJ66urk63AMgVtUqAGjSpiS3V6gGmwZVX9alPHyaHHj7EOECkqHzXEccOk4989CDjtupVqYkv1b6lJr5Uk8gCyK3Gxka7D7lgwQLdAiAbhBIIrL1790p3d7e88847ugVALqjVAdRgSYUSXq+0Yao7lnbJf8/bZhwgUlS+a8qC1+T2HAdxqtSksWofU5PIMvElkFvvvvuu3YfcvXu3bgGQDUIJAEDOqMGRM7GlWi3ANJjKRZkGhxRVqDJto7koNXms2tfUPgcAQKkglAAA5Ey+JraMLdPAkKIKVaZtNFel9jW1z6l9DwCAUkAogcB69dVXZf78+fKXv/xFtwDw0uqbbrIHR2oSPtPgKVc1tbZBzvvdEuPgkKLyXeOveUguueZW47aai2LiSyD3tmzZYvchX3rpJd0CIBuEEggsJroEcmfD0qX2oOiGw47K6UobpmKiS6qYKh8TXcaWmkz2uo8fbO+DTHwJeI+JLgFvEUogsNSVEvPmzeNKCcBjahCkJras/cTBeZnYMrYuu26unHdli3GASFH5rsrfPygX/77euK3msu6q18GgtS8y8SXgrc2bN9t9SHXFBIDsEUoAADxTqIktY8s0OKSoQpVpG81HqX1Q7YtqnySYAAAUK0IJAIBn7jnnHHsQVH/uJOMgKV9lGhhSVKHKtI3mq5j4EgBQ7AglEFj79u2TrVu3yq5du3QLgGysuuYae/Bz48mnGgdH+aqGB9bIlPnbjYNDisp3XbbwdZm79Hnjtpqv+uPg/okv1eSzALK3e/duuw+p/gsge4QSCCwmugS8U8iJLWOLiS6pYqpCTHQZW2qfdCa+7Fm1Su+1ADLFRJeAtwglEFhr1qyRCy+8UB5++GHdAiAT4YktrUFPISa2jK2zz/u5nPaDy40DRIrKd40cf62M+t6PjdtqPkvtm3ZwaO2roZ4evfcCyMSTTz5p9yGfeuop3QIgG4QSAICMuSe2nH/DfONgqBBlGhxSVKHKtI0Wopj4EgBQjAglAAAZm3/GGfYgZ07VZcZBUKHKNDCkqEKVaRstVDHxJQCg2BBKAAAy8vDFF9uDm5tPP8s4+ClkmQaGFFWoMm2jhSwmvgQAFBNCCQTWY489JkcffbTccccdugVAqtbMm2cPav507AkFn9gyto49/hQ5+ktnGAeHFJXv+sLw78kRnxti3FYLVUx8CWRn8eLFdh/y/vvv1y0AskEogcAilAAyU2wTW8YWoQRVTKVCiSOLLJRQxcSXQOYIJQBvEUoAAFKmJse72eqIqcFMMU1sGVumwSFFFapM22gxFBNfAgCKAaEEACBlxTqxZWyZBoYUVagybaPFUvXnTrL36XvOOUfv5QAA5BehBAAgJcU8sWVsmQaGFFWoMm2jxVR1J51q79urrrlG7+0AAOQPoQQCa82aNXLhhRfKww8/rFsAxFPME1vG1tnn/VxO+8HlxsEhReW7Ro6/VkZ978fGbbVYSu3TMw470t7HNyxdqvd6APE8+eSTdh/yqaee0i0AskEogcBqbW2VMqsDVldXp1sAmBT7xJax9alPHyaHHj7EOECkqHzXEccOk4989CDjtlpMpfZttY+rfV3t8wDia2xstPuQCxYs0C0AskEogcDat2+fbN26VXbt2qVbAMQqlYkt3dXwwBqZMn+7cYBIUfmuyxa+LnOXPm/cVoutmPgSSM3u3bvtPqT6L4DsEUoAAOJyJrace/F1xkFMsZZpcEhRhSrTNlqsxcSXAIB8I5QAABg9MGGCPTiZPeY84+ClmMs0MKSoQpVpGy3mYuJLAEA+EUogsF599VWZN2+e/OUvf9EtABzOxJYzhxxf9BNbxtZl182V865sMQ4OKSrfVfn7B+Xi39cbt9ViLSa+BBLbvHmz3YfcsmWLbgGQDUIJBBYTXQJmapI7NRhRk97d2/KccdBSzMVEl1QxValMdBlbTHwJxMdEl4C3CCUQWOpKifnz53OlBOCiJrdTgxAVStxVv9Q4WCn2mlrbIOf9bolxgEhR+a7x1zwkl1xzq3FbLfZi4kvATF0hofqQL730km4BkA1CCQBAmBp8qEFIqU1sGVumwSFFFapM22ip1Jyqy+xjgpr0FgCAXCCUAADYSnliy9gyDQwpqlBl2kZLqW76+ln2seHhiy/WRwsAALxDKIHA2rt3r3R3d8s777yjW4DgKuWJLWPrjqVd8t/zthkHhxSV75qy4DW5vQTnZnGXOiaoY4M6RqhjBRB07777rt2H3L17t24BkA1CCQQWE10C/XpWrbIHG2pSu1IPJFQx0SVVTFWqE13GFhNfAhFMdAl4i1ACgbXG6lRNnDhRli9frluA4An19IQntlSDDtNgpNTquz/6pXzth781DhApKt818sfT5Vvn/MS4rZZazb9hvn2suPnoo5n4EoHW0dFh9yGfeuop3QIgG4QSABBQalDhTGypZtk3DUJKtUyDQ4oqVJm20VItNQmuOmaoYwcAAF4glACAgPLTxJaxZRoYUlShyrSNlnKpY4Y6dqhjCAAA2SKUAIAAWnXNNfag4saTTzUOOkq9TANDiipUmbbRUi4mvgQAeIlQAoH16KOPymc/+1mZO3eubgGCYcPSpfZg4obDjvLFxJaxNeSLJ8nRJ5xuHBxSVL7ruK9+V4747OeN22op170tz8l1Hz/YPpaoyXKBILn33nvtPuT999+vWwBkg1ACgfXYY4/JoEGD5I477tAtgP+pWfPVxJa1nzjYNxNbxtZxJ/yHDPrySOMAkaLyXV8Yfo4cefSxxm211EsdQ+yA0zqmqElzgaBYvHix3YdsaWnRLQCyQSgBAAGhJrZUs+arQYTfJraMLdPgkKIKVaZt1C+ljiXqmKImvmRFDgBAJgglACAg5p9xhj14mFN1mXFw4acyDQwpqlBl2kb9VM7El/ecc44+2gAAkDpCCQAIAGeljT9/x38rbZjKNDCkqEKVaRv1W9WddKp9jFGT6AIAkA5CCQTWmjVr5Cc/+Yk8/PDDugXwJzU7vhosqNny/TixZWz917mT5LSxU42DQ4rKd51ROU3OPHu8cVv1U6ljy4zDjrSPNazIAb974okn7D5kR0eHbgGQDUIJBFZra6uUWZ2nuro63QL4j5oVXw0Saj9+sD1bvmkw4bf61KcPk0MPH2IcIFJUvuuIY4fJRz56kHFb9VupiS/VsUZNfKkm1QX8qrGx0e5DLliwQLcAyAahBAJr37598sorr8i7776rWwB/UbPhq8GBCiX8utKGqe78nxdkyoLXjANEisp3XbawV+54YI1xW/Vjzb9hvn3MUcceJr6EX+3Zs8fuQ6r/AsgeoQQA+JAaDKjZ8NXgwO8rbZjKNDikqEKVaRv1c829+Dr72MOKHACAVBBKAIAPqVnw1aCg/txJxkGD38s0MKSoQpVpG/V7OStyqEl2AQBIhFACgfXqq6/KnXfeKevWrdMtgD88fPHF9mDg5tPPMg4W/F5Tpt8u5/72PuPgkKLyXedf3Sq/ufrPxm3V7/XHwcfbx6LVN92kj06AP2zevNnuQ27ZskW3AMgGoQQCi4ku4UfOSht/OvaEQKy0YSomuqSKqYI00WVssSIH/IqJLgFvEUogsLZt2yYLFy6Uv/71r7oFKG1qtnvV+Q/SShumuvz6O+VHNUuNA0SKyneNv6ZNLp02x7itBqFYkQN+pK6QUH3Il156SbcAyAahBAD4gOrsB3GljXhlGhxSVKHKtI0GqViRAwCQCKEEAJS4oK+0YSrTwJCiClWmbTRoxYocAIB4CCUQWHv37pWXX35Z3nnnHd0ClKb5Z5xhd/aDutJGbM1d8rxc2viqcXBIUfmu/56/XW67/1njthq0YkUO+MW7775r9yHVfwFkj1ACgcVEl/AD1blXnfygrrRhKia6pIqpgjzRpamcFTnUKkFAqWKiS8BbhBIIrLVr18pPf/pTeeSRR3QLUFrUMnuqcz9zyPGBXWnDVN87f7J8fdyVxgEiReW7Kn5cK9/+/oXGbTWIpY5V6piljl2syIFS1dHRYfchn376ad0CIBuEEgBQgjYsXWp36tWs9gQSA8s0OKSoQpVpGw1yOStyqGMYK3IAAAglAKDEOCtt1H7iYFbaiFOmgSFFFapM22jQSx27VCjBUqEAAEIJACghatZ6Z+lPtcyeqbNPEUpQxVWmbZTaYa8WpI5lrMgBAMFGKIHAevTRR+Woo46S22+/XbcAxY2lP1OrwUNPlM8d/w3j4JCi8l3HDTtbDj/qGOO2Su2QOVWXhYMJoFTce++9dh/yvvvu0y0AskEogcBqb2+Xz3/+89LQ0KBbgOLmLP2pltUzde6p/vrCl/5TBn3lm8YBIkXlu4aOGCtHDTrOuK1S/cVSoSg1KoxQfciWlhbdAiAbhBIAUAJY+jO9Mg0OKapQZdpGqehiqVAACC5CCQAociz9mX6ZBoYUVagybaNUdLFUKAAEF6EEABQx1TlXnfQbDjuKQCKNMg0MKapQZdpGqYGlVuS4Ti8VqpY9BgAEA6EEAqurq0t+/OMfy7Jly3QLUFzUMnmqc67W82fpz9TrO+N+Jqd+/zLj4JCi8l1nnH+NfPO/Ko3bKjWw1LFOHfNYKhTF7IknnrD7kB0dHboFQDYIJRBYra2tUmYN+Orq6nQLUDxUZ9xZ+pNAIr361KcPk0MPH2IcIFJUvuuIY4fJRz56kHFbpczlLBWqjoGhnh59VASKR2Njo92HXLBggW4BkA1CCQTWvn37ZNu2bfLuu+/qFqA4qKU/bz76aLtTztKf6Vdj64ty2cJe4wCRovJdU+/aIQ3/s9a4rVLxywkm1FKh6pgIFJM9e/bYfUj1XwDZI5QAgCKiOt+qE64643Mvvs7YWaeSl2lwSFGFKtM2SiUvZ6lQggkA8DdCCQAoIvPPOMPuhKvOuKmTTqVWpoEhRRWqTNsolVo5wYRaFhkA4E+EEgisV155RRoaGuTFF1/ULUBhqU636nz/+TsEEtnUf197m4y7YrFxcEhR+a7zr/ofueiq2cZtlUqt/ji4f6lQggkUi02bNtl9yM2bN+sWANkglEBgMdEliokTSKh1+ln6M7tiokuqmIqJLrMvdUxUx0Z1jFx90036qAkUDhNdAt4ilEBgqQmK7rrrLlm/fr1uAQpjzbx5dmf7T8eeQCDhQV0xo1HOv+oB4wCRovJdP572sH31jmlbpVIvdWyccdiR9rFSHTOBQnrppZfsPuTLL7+sWwBkg1ACAArICSTUuvwEEt6VaXBIUYUq0zZKpV9qeWR1rCSYAAB/IZQAgALZsHRpOJBQnW1TJ5zKrEwDQ4oqVJm2USqzcoKJGw45RHrXrNFHUwBAKSOUQGCptaW3bNkiIZYZQwGozrTqVF//CQIJr+v2lufk0ju3GgeHFJXv+u9522TOff9r3FapzOqu+v5Al2AChbJr1y67D/nuu+/qFgDZIJRAYDHRJQrFCSRUp3r+DfONnW4q82KiS6qYiokuc1N31txCMIGCYaJLwFuEEgistWvXys9+9jN55JFHdAuQe+5AQnWqTZ1tKrs6p/JX8vVzf2ccIFJUvuubF1wvZ42dYNxWqezKCSZuO/FE6eOqR+TRU089Zfchn376ad0CIBuEEgCQJ6rTrDrPBBK5L9PgkKIKVaZtlPKm5lRdRjABACWOUAIA8sAdSKhOtKlzTXlXpoEhRRWqTNso5V3NHnMewQQAlDBCCQDIMXcgoTrPpk415W2ZBoYUVagybaOUt+UOJgAApYVQAoH16KOPyhFHHCG33367bgG8RyCR//r8F74in/3i14yDQ4rKdx37n/8lhx05yLitUt6WE0w8MGGCPgIDuXHPPffYfcj77rtPtwDIBqEEAqu9vV0GDx4sd955p24BvKc6x6qT/OfvEEjkq4Z+eZgcc+KZxgEiReW7vnjqD+Szxww1bquU91V30qkEE8g5FUaoPuSSJUt0C4BsEEoAQI44gcTNp59l7DxTuSvT4JCiClWmbZTKTd2/YrP8cfDxBBMAUEIIJQAgB5xAYuaQ4+1OsqnzTOWuTANDiipUmbZRKneljrnq2EswAQClgVACADxGIFH4Mg0MKapQZdpGqdwWwQQAlA5CCQRWV1eXjB8/Xtra2nQLkD0CicLXmB9WyannTDEODikq33X6j34vFd/5kXFbpXJbBBPIlccff9zuQz755JO6BUA2CCUQWK2trVJmdVTq6up0C5AdAoniqE99+jA59PAhxgEiReW7jjh2mHzkowcZt1Uq9+UOJlbfdJM+WgPZaWxstPuQCxYs0C0AskEogcDq6+uT7du3y+7du3ULkDkCieKpxgfXyWV3vWEcIFJUvmvq3X+TO1tfMG6rVH7KHUysmTdPH7WBzO3du9fuQ6r/AsgeoQQAZMkJJP507AkEEkVSpsEhRRWqTNsold8imACA4kUoAQBZCAcSxxFIFFOZBoYUVagybaNU/otgAgCKE6EEAuuVV16RuXPnyosvvqhbgPQQSBRnXTptjoy7/F7j4JCi8l0/qnlAfl1zs3FbpfJf97Y8RzCBrG3cuNHuQ27evFm3AMgGoQQCi4kukQ0CieItJrqkiqmY6LL4yn3FxDM336yP6kDqmOgS8BahBAJLTVDU1NQk69ev1y1AapjUsrjrt3+YJ+df3WocIFJUvuvH1y6XKdPnGrdVqnDlDibUMR1Ix0svvWT3Ibu7u3ULgGwQSgBAGggkSqNMg0OKKlSZtlGq8EUwAQDFgVACAFJEIFE6ZRoYUlShyrSNUsVRBBMAUHiEEgisPXv22BMUvf3227oFiI9AonTqtvv/P7nkzh7j4JCi8l2XNr4qty5+xritUsVRBBNI165du+w+pPovgOwRSiCwmOgSqegLheSec84hkCihYqJLqpiKiS5LowgmkA4mugS8RSiBwHrhhRdk0qRJsmLFCt0CRFOBxG0nnmh3Um8+/SwCiRKp74+vlm+cd5VxgEhR+a5v/mSGjP7BT43bKlVcFRtMqHMAYPLUU0/ZfcjOzk7dAiAbhBIAYOAOJP78nfOMHViqeMs0OKSoQpVpG6WKs1QwcdPXz7KP/eocQDABALlHKAEAMdyBxOwxBBKlWKaBIUUVqkzbKFXcpY79BBMAkB+EEgDg0rtmjdxwyCEEEiVepoEhRRWqTNsoVfzlDiZCPT36LAEA8BqhBAJLzSVx+OGHy2233aZbEHTuQOLOmluMnVSq+OuY474snx16mnFwSFH5riH/8R0pP+Jzxm2VKv6qP3eSfU5Q5wZ1jgCURYsW2X3IxYsX6xYA2SCUQGCtXLlSjj32WLnzzjt1C4Jsw9KlBBI+qS9+Zbh8/qRRxgEiReW7vnjaD+Vzg79o3Fap0ih1TiCYgNv9999v9yGXWn0HANkjlAAQeGvmzbM7nNd/4mCZf8N8Y6eUKq0yDQ4pqlBl2kap0ionmFClzhkAAO8QSgAItMenTbM7mbUfP1iaF7QbO6NU6ZVpYEhRhSrTNkqVXt1Vv9Q+VxBMAIC3CCUABJZah151Lm847CgCCZ+VaWBIUYUq0zZKlWapc8V1OphQ5xAAQPYIJRBYzz//vJx//vnS1tamWxAUanm3+WecYXcqZw453l6X3tT5pEqzRv/gpzLinEuNg0OKynd947yrZCQr+fiq7m15zj53OMEES4YGz+OPP273IZ988kndAiAbhBIIrNbWVimzOhR1dXW6BUGglnVTy7upzuTNp59FIOHD+tSnD5NDDx9iHCBSVL7riGOHyUc+epBxW6VKt9S5o+6kU+1ziTqnEEwES2Njo92HXLBggW4BkA1CCQRWX1+fvP7667Jnzx7dAr9zL/mp1p83dTSp0q95D/1Fpt7NLRxUcdTlTTul8cF1xm2VKv1S5xI75D76aFbmCJC9e/fafch9+/bpFgDZIJQAEAhqUjKW/AxOmQaHFFWoMm2jlH9qTtVl9rlFnWPU8tIAgPQQSgDwvfAKGyz5GZgyDQwpqlBl2kYpf5UKu52VOZ65+WZ99gEApIJQAoG1detWuf322+WFF17QLfAbdY8vK2wEry655v/JDy+/xzg4pKh8149+t0Sqf3eTcVul/FWszBEcGzdutPuQmzZt0i0AskEogcBiokt/c09oeePJpzKhZYCKiS6pYiomugxWqXONszIHE2D6FxNdAt4ilEBgbd++XZqbm2XDhg26BX7Rs2oVE1oGuH57w3yp/P2DxgEiReW7Lpj+iFxWe4dxW6X8WSqYcCbAVOciJsD0n5dfftnuQ3Z3d+sWANkglADgK+peXtURVMWElsEt0+CQogpVpm2U8n/Nvfi68PlITbYMADAjlADgC+75I9RkY8wfEewyDQwpqlBl2kapYNRd9UvDE2AyzwQAmBFKILB2795tT1T09ttv6xaUKnVprDN/hLqXl/kjgl1z7vtfubih2zg4pKh816WNW+X/3bvauK1Swah7W56LmmdCzXmE0vbOO+/Yfchdu3bpFgDZIJRAYDHRpT+oNeGd+SPqz51k7BBSwSomuqSKqZjoklIVO8+EmvsIpYuJLgFvEUogsNRSoL/4xS/k0Ucf1S0oNcsvucTu4NV+4mDmj6DCNfbHF8npP7raOECkqHzXmRf+Qcb8sMq4rVLBK3Wucm7neHzaNH02Q6l5+umn7T7k6tWrdQuAbBBKACg57uU+/3TsCcwfQQ0o0+CQogpVpm2UCm6pc9aMw460z2HzzziDZUMBBB6hBICS4r5dQ10Ky/wRlKlMA0OKKlSZtlEq2KXOXTd9/Sz7XKbOaercBgBBRSgBoCREra7B7RpUkjINDCmqUGXaRilKlVo21LmdQ92SCABBRCiBwFqxYoWUl5fLnDlzdAuKVezqGmomc1PnjqJUDTr2S3LU0FONg0OKyncNOWW0fObwzxq3VYpSpW7ncK/Ooc55KG6LFi2y+5CLFy/WLQCyQSiBwFq5cqUcd9xx9gzKKF5qIjDVUVPF6hpUKnX8iSPk8yd/2zhApKh81/FfGydHWwNO07ZKUU6p2znUOU6d69TtHM/cfLM+C6IYtbS02H3Ipdx2A3iCUAJAUVKTWaoJwFQHTV3aelf9UmNHjqJMZRocUlShyrSNUpSp5t8wP3w7B5NgAggKQgkARUd9QuRMZnnz6WcxmSWVdpkGhhRVqDJtoxQVr5gEE0DQEEoAKBqxV0eoT4xMHTaKSlamgSFFFapM2yhFJSv3JJj3nHMOV00A8C1CCQTW888/Lz/60Y/koYce0i0oJPfVETeefCpXR1AZ11ljJ8jw711iHBxSVL7rG+fVyBmjzzVuqxSVrNTEzuqcyFUTxWXVqlV2H/KJJ57QLQCyQSiBwGptbZUy6yRfV1enW1AIsVdHqE+GTB0zikq1PvXpw+TQw4cYB4gUle864thh8pGPHmTcVikq1Yq9akKdO1E4apJ01YdcsGCBbgGQDUIJBNbf//536e3tlT179ugW5Fvs3BEs9Ul5UfPb/iqXN+00DhApKt91RfObMu+hvxi3VYpKp2KvmmCFjsLZt2+f3YdU/wWQPUIJAHmn1mBXa7E7V0cwdwTldZkGhxRVqDJtoxSVabmvmlBXGqpzKgCUMkIJAHmjJulafskldkdK1ewx5zF3BJWTMg0MKapQZdpGKSqbcq/QoerxadOYCBNAySKUQGBt3bpV5syZI2vXrtUtyKW18+eHb9X4w2FHyV31S40dLYrKti7+fb38YGqzcXBIUfmu837XIr/67SzjtkpR2ZY6l6pzqjq33nz00UyEmScbNmyw+5CbNm3SLQCyQSiBwGKiy/xQl5WGJ7L8xMEyp+oyY8eKorwqJrqkiqmY6JLKdamrJtS51blqQp1zmQgzt5joEvAWoQQC67XXXpN77rlHNm7cqFvgpdhbNZjIkspXXfnHhTL+moeMA0SKynddcN0KmXp9g3FbpSgvyz0Rpipu6cid7u5uuw/ZQ/gDeIJQAoDn3KtqcKsGVYgyDQ4pqlBl2kYpKlelJo92bulQ52J1+yQAFDNCCQCeUfeyqntaVUdIzQyuZgg3dZiKojrelT/Xvyc/H39AvnS8dTC0joaqDjnmgAwfu1+uWrzL/LqE9abcdu0BOUR/r+HX7jY8J916S+5e9HepVr/nydbvp7932aEiX7p0n+H5+2S885ykdUAufSD29W/JvPr9MrrigHxGP+/Ikw/IhBv3yOKo58WW9W//7QH7+YeM/3uS5+a+TAPDUq+K8PuWvH4wz/w9Lquzvs+ZIp/+pH7uYSKDrMfn1ZufH657rOeMt577ucjrho4V+UWT4bnuuk/kVP2zvjvX8PWAlGkbTbse3y03/uF9+eFY61hwjH4frPrM8QfkzEnvyQ0PvmV+XcLaLZee5Xwv0/EgkwrJvDvekwnq93QdW8uOPCDfm/nugOfPs46Z4eckq0mGY96KvXLFJNffxPo5w8e/Lzcue3Pgc93V/nf5nnUcVf/u6sVJnluipW7pcFbpUOfmnlWr9NkaAIqLdfgGgmn37t32REVvvfWWbkGmVEfHmTdCVf25k4p8VQ1r4G53RhPXl365L63B9eI79stxrtdnHUqs2Cc/r4j+naLK1EHPKpSIDlXUYMcdhBxn/by4fw9XB9+bgU3mdeviZ+TiO14yDg5LubINJap+IvJxw3OdGvIzkatiXmPXfdbPPk4/75Mig04UOeqwyOMf3Gl4ja6Lf93/vMPGx/neAahL7uyR+nueNm6rKdeD78k39PuUqM6sTeeYE5IbrMF85PUe7Lst78n3rGOG+3dyl+mYmFUo0W4d75zgQwW1ww7IcUc6j61/T0u8sCESoh6T6Ljmg1LnYrXSlXN+VudqwonsvfPOO3YfUv0XQPaswzEQTEx0mT01kZY7jPjzd84rkXkjIgN3NfD+3qXvyVX1fTK9/j37igTnCgFV30ixk794wX45JSboyCqUsDr3o12fMn5mxIfyw5r3rN9R/54TP5RTqpOEEt99Xz8/Xu2TeY+7XrvsPTlTv/aH9ZErRSJhywH5+d2u54frTbnxN7qDb/yd8lt+negyHEp8W+TC6xPXZS3Rr73sikgg8cmviYyfI/Lbe0QuvkXkW9Zj+/tadep10a9T9YuJ+uvW8y5+INL+g5H97R8/2/perueHq0lkmL5K4rxkV1T4uDyZ6PKB92S4fo+O/JrrWHDjezLhu5EgUe2j4xtD5u8RVSF7n428rv+12YQSscfAI0d9ID+/9u/h31NdlWYKTSKhxAH5Ya37+GSoRe4rLazjTrX+eRXvy20dTnvk6o9Dxv1d7g4/31XWsW60/bt+KFcsM3zdh6XOzXUnReabeGDCBCbDzAITXQLesg7ZQDC9+OKL8stf/lIeffRR3YJUqY6M6tA4nRs1sVZpzRuxT37+3f0yPc7lzlFXPAzZLzcanhNVLe/JN3Rn/BuTPggPHjIPJfZI9Qj98w+1Bhn1Ay95jl+uUMJ4JUX8Cg8OBvybrU6+vmLjmN/sdbXrcgZMh34oVxVBB3/sBb+R03/0e+MAsZQrHEqMN389brVY26Vzu4YKFgzP+e4p+uvHiPwi5msV+rXDro5un1YXCTrG3xPzNauqrN9Tfe2onw38WpDqzAk3yHfG/cy4raZc1j42ftJ78ucVpk/+o69wKhv7nnkgHi7X89XxZWIkFMg4lAhfKWXV8R/KpXenfvubO5RI7+dbxzr9M0fPjPl51jG8/+9hHZNWuNrtUleI9L/uONPxzOelztXuyTAJJzLz9NNP233I1atX6xYA2bAOyQCQmtgwYtYppRZGpFohma4HVEk7yq7Lh+3bG1yfaGYaSrg76ePvSOVTT3dlHko4HfUyw5wQ4a+d9Z7Mi/papIN/ym/3uNoLW6bBYalXpqHEVdfp11n1jeuTP+dbt7i+Nk9kqG4fMCeE62sVs+J87ZMiF94X87UAlmkb9bb2SPUw5z3cLzcYn9Nf7iuf1PEl81DAqcgxIPEtE+bK+OeHj7VqToh4XxMZ3xjna4d+INPbY74WoFKTYc447EjCCQBFwTosA0BisWGEmtVbdWhMHR2/1J8vS6Gj7A4kfqIvE846lNgrE4b0v/6Qn/RlcK9zpqHEu3KFM+Gd4XXO/ddlFTGhxOL35RTVXmQdfNPAsNQr01DiFz/Rr7PqwpjbOsK1UORL+jmDfu1qnxW5GmLAPBXW9zpVf62iLvprF47tbx/i/l4BLtM26m2F5Kpx/X/zRKGEO5D44c39V2BlHUqseE9G268XOaUm/WAy45/f6FwNYXjd43+X7+nfafwd7q9FAudiClELWXfW3BJeqYNwAkChWIdlADDrXbNmQBihOjCmjo3fKvzJn9XhveJB03Nc9y2f9Z7c5rRnG0os2C/H6NePvyOTGeHzGUpEOvhZzZ+RgzINDEu9Mg0lztNzP5SdaL51w6nw9x/jmiMik1BirsgQ1X6MSFW8ECRgZdpGva3ILVZlh8a55cx9m5lrf802lLh75odZvT6voYQTog7ZLze459OhCCcAFJR1aAaC6ZFHHpHPfOYzMmfOHN0CR+xqGuo2Db9fGRFdrkuhh70fCRzC9a5M/4nuSB9vdW7dVwhkGUpEOugfylUPvit/vjF6eU57ubuES5a6QomYSrZ0YNq3byzQn7qmMu9GHmvQkOPlyC+MMA4OS7nCoUFMffI4kZPHi1QtNL/uW85SiV8Tuczwdad+4Ex46X5eBrdvjB/T33by71zPDXANPuUs+b/lRxm3Vc/KNQg37b9Ll/1dfui+zcz1tWxDifCxQR0HHt8tN1y7X850rdqjlllOdNyJ/PyYOlTkSxUfys9v3GOeIyN8rE319o135Sp9Bc+Zf0hnnp5glQon3Ld13HPOOazWYdDc3Gz3Ie+9917dAiAb1qEZCKZV1kl26NChMm+e1bOGbe38+XLbiSeGOyOlN4GlN7X45g/CHeqBlyO7ltFTgUTsxHNZhhKRKzREPuMsbRenzqzZa7i9I34o4a4zrx146XJ4cJDSRJeRttEznQ7+WzKvPnpAolYK+HlaE3VmXyecdKoMPvks4wCxlCteKOGuk6cmeF2qoYRVVa72dCa6vMpqO0y1HeeaMNP62g+swWB4GVHrvypEqTZMjunHOv7r58qgY08wbqveVGR5S1U/rI+ZhyZq3pu+AQP87EIJ9xUa1jHLmezSVIdGbhlxV9xQwlWHnGw41qpjXRoTXapbV+yr0Kzjdvj4tmKvXDretYzokf0BinlC0WCV+iDCPSGm+qBC9RHQr6Wlxe5DPvDAA7oFQDasQzCAIOsLheTxadPk5qOPDnc+1NKeQQwj7Grvkx/qOR3szmvUJb7Rs9YbJ3TLKpRw3UKhyuogj770Pblh0bty94pdcveDe2T6tR9ELbsXCQSc2icT1DKnasnABXv6X2fVbQv65IrqA3Kk67Xfuzlm8JLCkqATFvT/m8Md/BHOlSSusMaqI08+IF9yLWnqvlw8H2UaHJZ6fcsa6J96sciFt/Qv5+ks6XneRJFPO6trWBW7rGeqt33ECyWSLgnqut3D+R7h3+FOkS85v9thIoNOdP2u1r+nKiCTYJq2Uc+qJXLMUbeSRU9EG+c2M1dlF0pEggH7ZxxzQMZf+3e5saX/uHN3yz654tIPI6sZuY4hTs2r/VBOGbtfLr1xn/z5Qf0661h3w43vy/ecVYhUhY81TqWwJGh4JZJIeBI+7rluZ1HH2i8Ncx0fY6+AC3CpvsBNXz8r3D9QfYVnbr7Z7jsAgFesQy+AIHLmi7jhkEPsjkbtJw6W+nMn2WuZmzomwahIZ1Z1ngesfPGg1Ym1v2Z1bGdGBvxRdff7kVDit3t1e6oraLg+dSz7QKaHO9kx5RqEpHvrxOJF+xO8NnppQTtYcF31ELnsO7JkabiDf3f0jP7O97u7/gM9R8aHckUelws1DQz9XFfNidxKEbusZ7ahxLT7rO9xnP6aDhbCVz180nrdnf3PUyt42FdOnBKZu2L82f3P+/hI13wWD4h8V/+swybqNp+XaRv1pFxXQZiC0nm1OnD43AdylRMUxNSfw1dZHJDqu3V7u/lWi4HlujLrdNOtbv21eI5zHLAq6ZKl7npXrrK22/7fz3BFhPvfr4OF8FUPrr9H+Oq3cLCxKzwxaFRY02Ed25yrwqrTW73I76X6BrPHnCe1Hz/Y7jOovgPzTgDwinXYBRAk6vJL93wRNxx2lMy9+Dq5f8VmY0ckOBX9Sf+ZtYZP9l1XQaRX+xMu0RcpVygRu8pFVL3pus3DcD91wop+7aUtsV/vvwVjtDVodIcTE27cE75VJNzBd/2ON/5Gf88Bv3dkNZF8ToZpGhj6vapcgzcnKFCV7e0bdt0jcp71/d23YAwdK/KLpshzvntK/9fCk15arxmmv1/s6hy/vVp/n08afpYPy7SNZl+uENUagJuWD07l1ghjpTxJriuUSPiamMA1nUkm3cfdiX0Dv75ir1xhHbvdt2AMH/++3LjMCWgiIWr4b+RaMSR2QuHwxJ2HpnrcDlapvsKcqsuiJsXk1g4A2bIOu0AwPffcc3LuuefKgw9aPUafU1dFLL/kkvBVEapuPv2sgE1emaiiA4m4txrkM5SIN4O+rvAnoFZFL3mXvNwDlXRfa4cM+pPJyGXYiVbuiKzQkd5qIJnXt79/oQz/7sXGwaGf67Kp+u9slTsEyGqiyxRLhQz2VRJnRm7lSLhyx1yRQfG+5rP6+rm/k9PPGmfcVjMvVyBhuqpLV15DiQG3jkTXjb90vn+6t4m4fkbCsNZcKmSwQ9TvuK7QSLRyx+L39fK46f6ewavYeSdUH0P1NYJw9YSal0z1IZ944gndAiAb1mEXCKbW1lar01EmdXVW792H1P2esRNX3nD4UfYnHMG+RSO2QnLDLyO3LGQ994FnE11+kDjImPOBfl5+QwnntZF7tVVFwpQvGdb+D/+bkgxavKpPffowOfTwIcYBop8rXigRXhJ0RAZLgqZSLdZrdfDxrVtc7dbv4Pw+F8YuDRpn5Q4/1hHHDpOPfPQg47aaWe2WK77j/G3jBxKpVuR4kMkg3BWknm6es8Kp236f6c/JIpR43HqtfaVWzFwWd+zXv4vhqg3XMTyycgeVqFSfQt3+6dzaocq5esKvc080NjZa20iZLFiwQLcAyIZ1yAWC6e9//7vs2LFD9u7dq1v8YeMDD9j3eTodA1VcFRGvdkUFEsZbNtKtLEMJ9wAh0W0Z7mBhwgLzc+JV+FYLNTgYcPtGgnJ18KsXuy95jgxMhv9+4L8536HEgmUb5PLmN40DRD/XL36m/85WuW/fcLcPCAecWij602GRw6znG58Tp5ww5ONjRa5yf80VSgy4RSNAocQVi96S+W1/NW6raVfHHrnUCSTi3LKRbmUXSrhvB0t8W0bkeZEVMVIq160Wxts3ElQ4RI1dJtUVSgwIfwklsiq1pKjqczj9D2fuCdU38ZN9+/bZfci+vj7dAiAb1iEXQKmLnbRSlbrfU80VwVUR8Wq3XDUu0hk3LVWXUWUZSqhLh0/Rr48/0ZrrlohD07w/u3NfeI6H9CbJjCw7OKCDX2S3b6gyDQ79XuHbNGImurzq+shtFN+w/t/9mvBzrtOvtSp2/oeEdZ/IqXo1je/Ojfkat2+Ey7SNpl3t1r7r3GJz6AGpXpR9IKEqu1BihyyujyyhfOYf4h1H90j1MP27D1hFI3Hd/Qc9x0PC72+o9r/L9+zVNGJDVKu4fSPnpfoeqg/innvCub1D9VkAwM067AIoRc48Ee6lPNWkleoSyuYF7cZOAqWrfa9UuyaIm7DAm869XUlDCdflzlYNvH3CGuC7Ls02LTsaWaJT5Ljf7HV9bbdc8aMP5ef1u2NCA6ei5844xXCrRdxydfBNHXUmusxtnfc9kTHXx1yJ4Cr3JJdDfh3zddftFfaynu6v6XImqYwNNJLVxdbPUq87zPr5A343JroMl2kbTauW/V1+6KwycfyHcpVpOeIMK3ko4bp9wvgcd9D5gWEpzehVfaKWMbaOZcMnvSc3Louz2od72U4VwKa8TGckRD1mkrNqkKuY6DKvpfoksbd3qL4LAQUAh3XYBYKpp6dHbr31Vlm7dq1uKX7q8sfYIEIt5amW6eL2jNTLfevDmb/9u0yv70tSe1Jfwi7rUMKqqI74AflezT65TS3T9+AemV7zgXwp7lr6ru9tPUet/V/9h/5/w1XX7pfR7jX/rdfemPIVFm/Kjb/RHfx4V2+4lgSNhDyFWRL0N1f/WX5wWZNxcFiqFZ6E0hrEDxkj8t3fiVx4vbX9TLUG/k6goOo4kV8YbtG47IrIVQuftL7X+Dkiv71H5OJbRCqcT7CtGnb1wNfGrSbr+foqifNcq3C4K7wkqHueioAtCXrulffL5CvqjNtqqhW59eGAjNf7dMJqjBdMDqzsQ4nooFStfqFW65mnjlkt++SK6gPyGf21qOU3VYVvoxD5zPHWse7S9+Qq+9/wnlSPPyBHOsc6q76Rzu11y96T0fZr4x13XEuCuufHYUnQnJfqq9hLi1p9l9iAYuvjj+veTvHbsGGD3YfcuHGjbgGQDeuwCwRTKUx06UxWGXtrBkFEduUOJVKrND4x8yKUsEp18sPhg6EOOfkDmR5e8s4p120UCeqQk61/z4o0Pml1/k2HfihXxQ0Woq/CUMuIfsn5ZNeqrCcQTaP8ONFleLLKBPXxL4tU3WN+vaqqn0SCCVMNsr4e70oMUzlXZxyVaA6KO0W+pIMLtYzooBNFPu08Ps76HvcZXuOz8mKiy0gokWKlMSGkF6GEfWVC7Yfh8MFUn/nO+/Ln2CsdwrdRJK4zr03jqi77WNT/uugryWLKHf4eaR2vhrlCkAGBL5WLCgcUriso3HNQFPMkmUx0CXjLOvQCwfTaa6/JvffeW3Qpt7qU8Zmbb7ZnrnZO0vaJWt+acVf9UuPJnUq9SiGUsEutv1/9oQx3De7VYH/8H/bJ3R2G59sVknkL+j9l/NLJEunwW53tL33nA6m+MY2rPuyKdPCT3+4Rkttu3C9nun7ukV9Tt5N4NF9HivW7mXfJ+GuWGQeIpVyX3SLy3bHWwP7LrnDBGuAPOtNqn5baihmX1YlUWM8PBwP69efVm58ft5yJKq3XX5gsWGjq/72POkz/TOu/J48XqU4QoPipflL7qFw+407jtppqFX8ooevBvv5jj3O7kFXHVSS6pcyqZXtkurqSq+KAHHdk5HXqyonR1e/JDQ/GubUjXoVD1BRu91i2z/59wz/3yANy5qT35M/phLaUJ6UCCtXHUX0dd99H9YVUn6jYbvPo7u62+5DqqlsA2bMOwQAKSa3n7VwN4b4tQ9Wfjj3BniiKOSIoKr0yDQ4pqlBl2kYpijKX6vOo5ctVH8jdJ1J9JNVX8vNSo0BQEUoAeaZOpM7cELedeGLUCfeGw4+yL2VUS2rdv2Kz8WRNUVTyMg0MKapQZdpGKYpKXqovpPpEsbd5qFJ9KNWXKvZbPQAkRyiBwHr33Xflr3/9q7z55pu6JTfcV0LEhhBqbgi1njdXQ1CUd/X/7l0tv5m7xTg4pKh818UN3fLnRU8Zt1WKotIr1VdSfaYbTz41qj+lygkpVJ9L9b1yKRQK2X1I9V8A2SOUQGDlaqJLNXv049OmyT3nnDPgdgwVQqgTqboskbkhKCo35ceJLqnSLS8muqQoylyqL6X6VKaQQvXBVF9MzUnh9coeTHQJeItQAoH14osvyq9+9St57LHHdEv61ElOnexMV0GoUrdj3PR1roSgqHzWD35ysZxx/jXGASJF5btGTfij/Ne5PzduqxRFeVsqpFB9LtX3ir3dQ5Xqq6k+W7ZBRWdnp92HfOaZZ3QLgGwQSgApUPcqOldAxAsgVM065VR79mg1i/S9Lc8ZT5gUReW+TINDiipUmbZRiqJyX6ov5qzsYbqaQpUTVKg+nurrMT8FkH+EEoCLEz6oBF3dl6iWolJrZptOYk4AoSZg4ioIiiquMg0MKapQZdpGKYoqTKk+m+q7JQoqVN9P9QFVX9C5qoKwAsgdQgkEjhM8qImQnCsfEoUPzi0Y6p5FroCgqNIo08CQogpVpm2UoqjiKeeKCtXXU30+1fcz9gl1WOFcWaH6kgQWQPYIJeBLvWvW2CcJdcJQpU4gqtwnlgus+phV39GP/3TcCVHhAxNRUlRp1tGDvyhHHvdV4+CQovJdg0/+tnz6M0cYt1WKooq7VF9QXVXhhBWqr6j6jD/Qfchxug/plNPfVH1P5woL1ScFkBihBEqOOsC7r3SIFzqYat7pp8ui731PVl1zjdxxxRVy9JGflV/9eprxRERRVGnWCSefJkNOGW0cIFJUvuuEr58nxxz3ZeO2SlFUadaUy/9k9yFnVlXZfUrVv1Rl6nu6y+mvqttCVP/VudJCFRBkhBIoCs4BWZVKltWB2rmtQlW8WytiyzkpPHzxxfZJYsPSpdKzapX+KQNteHWP8WRDUVRpl2lwSFGFKtM2SlFUaZfqQ5qoWzlU31P1QVVfVPVJVd90zle+Yuy7mio2vHCuulDFlRfwI0IJeMYdLKhywgVVzqSRTqm1o00H4XilDuRO4KAO8KrWzJtnH/QThQ7JEEpQlD/LNDCkqEKVaRulKKq0K14okQqn/6r6sk6/1unnphNeqFKrh7j72E7fW5X7SgxVBBooVoQSAeNM8mgqd4gQW+6rFlTFWxIz1frDv/1b+OCryrmyQZVzdUM2YUOqCCUoyp9lGhhSVKHKtI1SFFXalU0okQrnigtVzlUXqpwrL5xSfWpTXzvVig017jnnHONYQFVsyOFUqKdH/9ZAZgglPJZo0J9KmQ4Apoq98iC2sg0NUin3AVGVc7B0yn0lQz4ChnQ999xzctZ3vi81f7rLeLKhKKo069vn/ES+evZFxsEhReW7vj7uSjn92z8wbqsURZVmTa9fYvchH7f67sXG3fd2X4mhKjbQUJVtqJGonNVKUi1nVZN45b6NJd1CcSuKUELtJKYNM5tKdQ4CP5f7lgennEkeTRUbIvz/7d1NiBtnnsfxzgubZMDJkFftLoPTg7NshxziwwzWMgtjgZlNz2bBvTS7NoSMIwzbaHwIwgev8KURPpgmCY3IoRENmaA+hKgXGuSDQX0JqA8G+ZAgHxzUB0N0CEQHH3Tw4b9VUr08Vpekqu5WP6qnvh/4H6xWd4IePU/V86unnlLLxAR0Z2dH5qzP6S/X2OiSokyqV9/4e3ntH94JnCBS1EnXP/7T7+XvXngp8LtKUVQ866//+0X/HPKrr75yzirNoa7QGK69zz8PnEP838cfH5hz2PX56dOBc5QkVdA89ahlf+ammYlQ4n+mkNDlfvUrLXX11ClZfuONQ9cfrc4bpv7lt7+Vd999lzpCLSwsyO/T/yr/9h//JX++eJmiKEPqwp//U/7wp/+WP/77xxSlvf7wp0v972TQd5WiqHjWB9a5o30OaZ9LBp1jUoer373zTuC8x60Lv/lN4PwpTAXN206iCs8+GzhXPUr95de/dmbR5piJUOKll16SZ555Rt58803K0HrDGgzsNn7hhRcCf06ZUa+88kq/nU+dOhX4c8qMcsfsV199NfDnlBn1/PPPy7PWyVTQzygz6vXXX+/35RdffDHw55QZ9fLLL3NsTkDZ/dhu59deey3w55QZ9dxzz/WPz6aZiVDi7Nmz/ckMzPXLL7/0l7ldvHjReQUm+vbbb/vtvL6+7rwCE+VyuX47//DDD84rMJF9BfCtt95y/gUT/fTTT/2+fOnSJecVmOjrr7/ut/PGxobzCkz0ySef9Nv5xx9/dF6Bid5++205ffq08y9zEErgRBBKJAOhRDIQSiQDoYT5CCWSgVAiGQglkoFQYooIJcxHKJEMhBLJQCiRDIQS5iOUSAZCiWQglEgGQokpIpQwH6FEMhBKJAOhRDIQSpiPUCIZCCWSgVAiGQglpohQwnyEEslAKJEMhBLJQChhPkKJZCCUSAZCiWQglJgiQgnzEUokA6FEMhBKJAOhhPkIJZKBUCIZCCWSgVBiigglzEcokQyEEslAKJEMhBLmI5RIBkKJZCCUSAZCiSkilDAfoUQyEEokA6FEMhBKmI9QIhkIJZKBUCIZCCWmiFDCfIQSyUAokQyEEslAKGE+QolkIJRIBkKJZCCUmKIvvviiXzCXHUrQzuazQwm7jQklzOb2ZUIJszFmm88OJWhn89mhhN3GhBJmc/syoYTZTB2zZyKUAAAAAAAAyUMoAQAAAAAAtCCUAAAAAAAAWhBKAAAAAAAALTSHEm2pfJTq7xQ7N7cklUfOywCAmdLdGTxxY2614bwCYPZ0pbZin1PNSXHPeQkAEAvtrcuS6s+L52Rpq+O8mgwaQ4meNFbTTiBBKGGEXkea22UprCxJ5j03bLLqTFoyV4pS2UtW5zJV71FTqpsFyS1nZCHl9t95SV/ISnGrIZ0nzhthkJ7Urzt9mlAidroPa1KyxuX0Gae/phYks5yT0i5jsnEe1yXvjMuEEmbpPqhLZT0v2Qtpme8fd5W+fKctXed9iLEnXWntVqR0PSuZc/ODNqadE6O3V5S02+ZWEUqciJ40P1v0kqBBEUrE2r01/yA5puavVKTNpDW2Ot9cDmzXp+pMVqr7zi8g9nqdltRuKeM1oUSstL/Jjh2b09dq0nbeixh70pPOg5oUP/QvCBBKmKN5S7nQM6JS54vSeOz8AuKnU5Wsd6FndPXbmWTCOL37a7I41P6EEifAW5qSuizFm0vOh08oEWt7RacdU7LwYU4Km1Wp79atqkr5ZlYySkdL32pKz/k1xEtny+mvqQVZXClIedtu47rUttYke145aTpXlCaNHGteWw8XoURs9O75V11S5/NS3mtLt9u1Jq91KX+a8YKmxQ1iifjqSGVZ6Z9KEUqYo7HqtOuZtGSvl6RyZ3DsrW7mZcldAWVV6qOq9Y1ALD2qyJLTjvPnspJfr0itfx5d66+QoZ0Ntl+Ry/15Ukourxa87wGhxJT5S1PS1gGzp5z4EkrE2t6aLK1Wpfmz8+9hXoezKlUUpjXx1NkuSnG7NWL5oLpHzJzkdojy46yzk5fMhYxX3q06hBIx0ZbyB06bvV8IuIKq3JIzl5XqqLEbM64jtU/9fpo5v+CFTYQS5mh+mR19u9XjhhTPOX197qys3XdeR7w8qkr+aknqj0Zc0XmqnZkzGUNp17R1ftVTwilCiSlSA4nC3cGE5SihRPdhXSqrQ/ddjamznzWd34QOjZtuWzCYGuthWTJun2PyahDlauxR2rXTlOp6TpaUidPYWqlx/+xhKX1xcXPESgjlPdltPmkjKCe0Rwol6Kux0rub99oiaROZJGn/zV/BSOhoADWQuFEfjKFHCSViPm6fXCihLk3Z8k+QDhVKdFtSvhryA/cqI+WHzu9DCy+UiLRSoivt3YoUr2T8TdrGFlcJtHpck5zbFlEmr0/sTVJLQ5tnjquc1DgDPkFHDCWs9q2tLobad0at/F3uATosf/+Xcce+phTd/na9Hvq2Oi4IzLCjhhL01Xh6UPJu1Yo0keHYGyt+KBHt4h5j9izyVxenPqr4ezsdJpQwZNw+mVDi57oU1KUpzsu2qKFE70FZsu7k9MySFLfq0up0+/fJtvfKklfua09/1ui/PigOmFopO4KnQw523e+ttlaf4hGmLpTZtE0nZTBN3QrXzp07RVkMFTgpFWECheNwhFDCHv/dcTmVkex6VZr7g3HZ3phvbVk5SVqpSscbs7u08RH4G+PlpT7yg1Ta9f01mdhjuSAw+44SStBX48vb12tOLn8TbiLDsTdu/FvyUmGvcDNmz6iu1G84T6A8N7RBbdRQwqBxe/qhxNC9MsOdKFIooexLkL5WDX6Kg3qllo1g9Ot1pXWnJFnnOzB/xWo350ej9aS16e8YP79clMpuy+lIbWls5pWNM9Oytud3sC47T2ulLiGdnMBag/JNd7O9lGSulqR6f7ARX9d+4sPtJSX1zUn1kdLOnBWdsEOGEsr4n/rQmvQGnkW1pHTB+dup4uSJMULoSm3F+Uznxq9M8zbQm/A+LgjExGFDCfpqrLU3Mk7/CzOh5NgbK0961hhbkYLzZJ2wT1lhzJ5VPeu4qwQSw2NtlFDCsHF7yqHEiKUpivChhJIQjvhbA8rJ83KFUEIT/0R3UPMfhH++svd0lrm05LaDW7p3J+f97bBXBTBtysZ6qcKYq7M2ZVBOLcraveBvRutL90QrJcV7zovQ4DChhLKR4vCVgCFhJ8YIS2mv0KHEmGMwFwTi41ChBH011pSVqHMfTFotyrE3Lvz50aBS712W4lZDOmEeq8+YPbPUJ1BWgh6fHzqUMG/cDgwlhieUkUo5YfXuaf3nvNScpSTD1drwQ4ny987rAR+s3zkXJ6TAbalcdP5fCCW0OfgdmpfFlZL1PZiQwD4sy2L//U/vPXJA1OVNmDr/Xker7SYERfaKCjd4sp/CM5KyJJVNnUYbPnmJVKHGyeihhL2x8dn+fyM1cdWMvwkuE53jET6cnxxKcEEgVg4RStBX46xntclZp02s4+m98e3HsTc+Dh7XU7JwqSiVEUGSjzF7ZnWqcrnflguS33FWJw3X92V/frPRcl4/2FdNHLenG0oc9kT5wElvU9beH/wsdWPC/Wy9uuSdv8NGLeM0pOh+3oeoiQepx07n6rSkcUddJjbuQKgMkJPS/n1CiVCUE9ToFW6fF5v/ZJ1JB0FLz/ruuf150v2p3xW8/x9OjEabvVCiK9UrzvsnXrmzTqDcpYXs4H9Mji+U4IJAzEQOJeirceavLJ07sGfbARx746XnnEd329LcrSp7A4y/aMeYPcMOfU4+HCaYOW4H377hTigPU8oqh2MLJawJz2DQTU0cHHu7Bee9PIVhvCmHEgf4t/KMXNp/f81L/Qq7Yw+Xyt4FtPNYJxFKKMsEw9zr2N3OOn9/0gHT+tZ498my4/dY3snLYWp8XxuIGEoo37uJoaG3OorHUh4fpb0mXBkZH0pwQSB2ooYS9NXYUi8GhNmvi2Nv3Cm33oxsQ8bsmXZcoYSh4/b0N7qcwA8uRk+A/F3EC2NPruwO691fMzE5Qi9wghKuemHuaRumJO9BS40aq25oMWl5kdXON5z3htkxPtF6ge0XtiZOV5VAYtI9bQPhr+D20113jwrrhIvzIp2ihRL+4yitcT3onklFe2Nx8N6J+5AgvGPa6JILAvETMZSgr8ZTpNWJfRx7jaCsEs5sBLQ6Y3b8hQgcTB23YxBKRBhIlf0IcjsMozNn7HO0lefl35ww6VE2dVrcJHrSZr/q7+wcKpCwePfTWd+Bv01oO6U/T1o5g2mLFkqMneiqDvGoYIQTLsxX2jUg4OWCQAxFDCXoq/FjBxLuE8jCBRIWjr1mUDenDDgWM2YbIEQoYeq4HYNQQpmsjg0l2t7JVehBGifLOpAO2jqgo0VYiuSnfjmp/ey8iJO1X/ECibCPp+pTvgPjT5jVXYVZDaNflFCiI9WPnPeOPWAqG7SFDbUQWnfHfULRuEcEKsfXA/eackEgliKFEvTVuOl+5wcS4R6x7uDYawalfx88FjNmG2HifMjccTteKyXse9uCPtQnbalccTaAGfWIFWjXvO3uEB3wiKl77pKzCaGEt3RtwtM5MDW9+2uy6F2lKUsrwkDnLzkbf2LU+67g7C8yYYdwnJAooYSa4mek9MB58Sldadxyn5NPG0+Fd9I5ZkWZ9Z6M856D4y4XBGIpUihBX42T9k7OW22avlGPdFsFx14z+PuCBD0OnzHbCCEu0po6bsdiT4nWuruxi31VNi/lPecxKvtNqW0VZcldQn4mK1UCCQ06Uv3UbpeOjPra2+m+ezANTN/V9PdaLfjvqFfnGUi16N4t+CdFVjtFbgPlak3my5bz4tPUK0ETdxPHCYkWSnTv5LyQce7MkqzdaUnHHrPtp/FslyTrPo0nleHEd2qUNnu/EHCVRLkiOmcdOw+sOuOCQCxFDCXoq3HQk9aG+5SNlCx+1ox+XOTYO/Matwf9rztqzzZ1D6/AlcKM2UYIEUqYOm7HIpSQxw0pnnM+/MBKSebTirRYgaSJMhCeSUv2+pqUt+tS37Vquyz5Swt+5xmZ2LWk5LWx3Z5laewPNlts369JZXVJ5p2/EWnJIo6R8sSWD4tStdt3QjUfDbV1tyY596A6Ny9Lt2vS6tjt3JHWXlVKV91k1/oO3OKkSCd/bB5fwRMf68THfdrOiJpfXpP66Es5OAa9e8pmeEqg33lQl/I1P+wfda8pFwTiQJ2IjKtRy3zpqzNPmaQsXC0HHmuHqzU8YeXYO/O8q9+pBVlcKUhpqzZozzsVWVtZ9M6B7TYatVKYMdsAIUIJU8fteIQStsdtqa3nZOmc+5xe6wM/l5HsakUawxMfnLCOVK+M7xz9SmWkcHd0cqTuJh1Y1u/nt1qRliziOEV/jGzQgKo+Vz2w7NR3lzNg3Y4WSti60rROgrLn/VAy9V5GllZKUntALz4p7W+yysnswRob8nJBIAaOGkrY6KszTV1JGrKCxmWOvbOteduf34yueclutkaHRozZ8RcqlLCZN25rDyVgip50H9Slsp6X7IWMLHiJfEoWzmclv1mTdog+0tuvSWllSdJummsNwOkLWSluNaRD9qTZ8YQStu69ihSvKN+T1IJklnNSGrd0EcChdB8GjKt2fwszAeGCAKDXMYUSNo69s62335TqZkFyyxllvLbH3CXJ3bbG3DCZEWM2YopQAgAAAAAAaEEoAQAAAAAAtCCUAAAAAAAAWhBKAAAAAAAALQglAAAAAACAFoQSAAAAAABAC0IJAAAAAACgBaEEAAAAAADQglACAAAAAABoQSgBAAAAAAC0IJQAAAAAAABaEEoAAAAAAAAtCCUAAAAAAIAWhBIAAAAAAEALQgkAAAAAAKAFoQQAAAAAANCCUAIAAAAAAGhBKAEAAAAAALQglAAAAAAAAFoQSgAAAAAAAC0IJQAAAAAAgBaEEgAAAAAAQAtCCQAAAAAAoAWhBAAAAAAA0IJQAgAAAAAAaEEoAQAAAAAAtCCUAAAAAAAAWhBKAAAAAAAALQglAAAAAACAFoQSAAAAAABAC0IJAAAAAACgBaEEAAAAAADQglACAAAAAABoQSgBAAAAAAC0IJQAAAAAAABaEEoAAAAAAAAtCCUAAAAAAIAWhBIAAAAAAEALQgkAAAAAAKAFoQQAAAAAANCCUAIAAAAAAGhBKAEAAAAAALQglAAAAAAAABqI/D8Pb5RXz3nwOwAAAABJRU5ErkJggg==\n", "text/plain": [ "" ] }, "execution_count": 187, "metadata": { "image/png": { "height": 250, "width": 500 } }, "output_type": "execute_result" } ], "source": [ "Image(filename= \"IQR.PNG\", width = 500, height = 250)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Based on the IQR, a rule can be defined to identify outlying observations. For example, an outlier is a point which falls below $\\text{Q1} - 1.5\\times \\text{IQR}$ or above $\\text{Q3} + 1.5 \\times \\text{IQR}$. The range in this example is wider, as the 95th and 5th percentiles are used, but the same idea applies." ] }, { "cell_type": "code", "execution_count": 138, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "False 432\n", "True 10\n", "Name: pred_outlier, dtype: int64" ] }, "execution_count": 138, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df = df.assign(IQR = df.pred_upper - df.pred_lower)\n", "df = df.assign(pred_outlier = (df.charges < df.pred_lower - 1.5*df.IQR) | (df.charges > df.pred_upper + 1.5*df.IQR))\n", "df['pred_outlier'].value_counts()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Out of 442 total patients in the holdout set, there are only 10 patients who have observed claim costs which are outside of the 5th and 95th prediction intervals. " ] }, { "cell_type": "code", "execution_count": 139, "metadata": { "hideCode": true }, "outputs": [ { "data": { "text/html": [ " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
agesexbmichildrensmokerregion
220235403
12027301
461491002
1408211
340538502
360539012
16169200
460467000
281146503
80245400
" ], "text/plain": [ "" ] }, "execution_count": 139, "metadata": {}, "output_type": "execute_result" } ], "source": [ "style_table(df.where(df['pred_outlier'] == 1).dropna().head(10)[['age', 'sex', 'bmi', 'children', 'smoker', 'region']])" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "A simple check confirms that these are, intuitively, atypical patients.\n", "\n", "1. Patient BMI is extremely high.\n", "2. Patient is listed as 12 years old and having 3 children. This is likely a coding error.\n", "3. Impossibly high BMI.\n", "4. Impossibly low BMI.\n", "5. Impossibly high BMI.\n", "6. Impossibly high BMI.\n", "7. etc, etc.\n", "\n", "Depending on the data, there could be many reasons for there being values such as this. These could be due to recording errors, database errors, outlying values due to forgery or tampering, or errors in way that the data is retrieved. \n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Discussion\n", "\n", "**Percentiles cannot be aggregated like averages.**\n", "\n", "While this is true for averages, some algebra will show why this is the so. \n", "\n", "Take a given vector $Y$ with $n$ observations.\n", "\n", "$$\\text{Mean}(Y) = \\frac{1}{n}\\sum_{1}^{n} Y_i$$\n", "\n", "Imagine that there is some binary variable such as Gender, called $j \\in \\{A,B\\}$, so that each $i$ is in exactly one group. Then converting from group averages to the overall average is straight forward because the sum breaks apart.\n", "\n", "$$\\frac{1}{n}\\sum_{1}^{n} Y_i = \\frac{1}{n} (\\sum_{j = A} Y_i + \\sum_{j = B} Y_i)$$\n", "\n", "This means that average predictions can be sliced across any dimension of the data. For percentiles, however, this is not the case because there is no such algebra. This means that a graphs or calculations which use an \"average\" or \"sum\" of the quantiles will be incorrect.\n", "\n", "
\n", "\n", "**Training time can be slow.**\n", "\n", "In this example, the data was relatively small and so training the model took only minutes. In larger data sets, this training time is often an issue. If the 75th or 71st percentiles were needed, then a new model would need to be trained. \n", "\n", "**Outlier detection is not a substitute for data quality.**\n", "\n", "The method of outlier detection relies on learning from good examples in order to detect outliers. There is nothing that a model can do to improve the situation if the entire data set is of low quality. Just as with any type of model, the accuracy improves with a larger sample size and more features. The prediction intervals will be more accurate when the data is better.\n", "\n", "## Further Research\n", "\n", "The output of a GBM in a classification setting is a probability, and so in theory prediction intervals can be made to provide an upper and lower bound to this probability; however, in [scikit learn](https://scikit-learn.org/stable/) there is no support for quantile regression for classification problems today. This may be added in the future.\n", "\n", "While this example used Python, quantile regression can be performed in most software packages that allow for custom loss functions. In R, for instance, this can be done by manually specifying the score function in XGboost. An explanation of how this can be done is available from [Benoit Descamps](https://towardsdatascience.com/regression-prediction-intervals-with-xgboost-428e0a018b).\n", "\n", "Quantile regression also works for decision trees, random forest, and neural nets. [Max Ghenis](https://towardsdatascience.com/quantile-regression-from-linear-models-to-trees-to-deep-learning-af3738b527c3) published a tutorial in 2018 which describes several of these implementations.\n" ] } ], "metadata": { "celltoolbar": "Hide code", "hide_input": false, "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.7.1" }, "toc": { "base_numbering": 1, "nav_menu": {}, "number_sections": true, "sideBar": true, "skip_h1_title": false, "title_cell": "Table of Contents", "title_sidebar": "Contents", "toc_cell": true, "toc_position": {}, "toc_section_display": true, "toc_window_display": false } }, "nbformat": 4, "nbformat_minor": 2 }