{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "[Table of Contents](00.00-Learning-ML.ipynb#Table-of-Contents) • [← *Chapter 2.03 - k-Nearest Neighbours*](02.03-k-Nearest-Neighbours.ipynb) • [*Chapter 2.05 - Logistic Regression* →](02.05-Logistic-Regression.ipynb)\n", "\n", "---\n", "\n", "# Chapter 2.04 - Decision Trees\n", "\n", "Decision trees are a family of related algorithms which build logical trees to represent data. Once constructed, a tree resembles a flow chart, starting with a single node or decision point, with two or more branches as optional paths. At each decision point, some feature is considered, and the correct branch is selected based on the value of the feature in the context of the logic presented.\n", "\n", "For example, the starting node may consider feature $F_1$, with two branches: $F_1 < 4$ and $F_1 >= 4$. If the observation's $F_1$ value is 2, then the first branch is followed. If the observation's $F_1$ value is 27 then the second branch, and so on.\n", "\n", "Depending on the branch taken, another decision point may be reached, most likely for a different feature. Sub-branches will typically consider different features even at similar depths (that is, the number of decisions made), but will vary based on the parameters specified for the model to decide how and when to create a new branch.\n", "\n", "When the model stops branching, instead of a decision point we have a node (called a leaf) which contains the predicted label for classification. We can take this value as our classifier output for a given new observation.\n", "\n", "Decision trees allow us to consider complex relationships between features (unlike Naive Bayes - which assumes each feature is independent), while allowing us to train a model (producing a decision tree from data) for far quicker predictions than achievable with the k-Nearest Neighbours model! \n", "\n", "As a logical structure, decision trees are also straightforward to inspect, reason about and explain! The trade off is that this simplicity is prone to over-fitting - when we create a model that is able to reproduce its training predictions with low error, but has a high error rate on unseen data. In this chapter, we will see one method for resolving this for a given tree.\n", "\n", "## Forming a tree\n", "\n", "Building a decision tree on given dataset follows a simple, iterative (actually recursive) process. As with the previous classifiers, we start with a dataset which contains multiple features (X's) and set of labels (Y). To build a tree, we want to find the best X to select as the root node to create a split. There are a few strategies for making this selection, but we will focus on one that uses *Gini Impurity*, measures the misclassification of splitting on a split. We want to select the split that minimises misclassification.\n", "\n", "Let's revisit an example from when we introduced Naive Bayes classifiers:\n", "\n", "| Employment | Default | Count |\n", "|---|---|---|\n", "| FT | N | 59 |\n", "| FT | Y | 1 |\n", "| PT | N | 36 |\n", "| PT | Y | 4 |\n", "\n", "To calculate the best split, we first calculate the Gini Index of the whole data:\n", "\n", "$$ Gini\\ Index\\ (GI) = 1 - Prop(value_1)^2 - Prop(value_2)^2 - ... - Prop(value_n)^2 $$\n", "\n", "where *Prop* measures the number of observations with the specified target value as a proportion of the total remaining observations.\n", "\n", "In this example, we can calculate the base Gini Index of *Default* as $ Gini\\ Index = 1 - 0.05^2 - 0.95^2 $\n", "\n", "To calculate Gini Impurity, we use a combination of Gini Index calculations:\n", "\n", "$$ Impurity = GI(Target) - Prop(split_{left}) \\times GI(Target \\mid split_{left}) - Prop(split_{right}) \\times (Target \\mid split_{right}) $$\n", "\n", "For a split *s* on *Employment*, where the left node is *FT* and the right node is *PT*, the Gini Impurity is calculated as:\n", "\n", "$$ Impurity = 0.095 - 0.4 \\times (1 - \\frac{36}{40}^2 - \\frac{4}{40}^2) - 0.6 \\times (1 - \\frac{59}{60}^2 - \\frac{1}{60}^2) $$\n", "\n", "Impurity is calculated like this for all possible splits, and the split resulting is the smallest value is selected as the root node, producing a left and a right branch each with a corresponding subset of the original data. To complete our decision tree, we repeat the same calculations and continue to select splits on smaller and smaller subsets until reaching a stopping condition (such as reaching a maximum tree depth, or when $GI(Target) = 0$ (meaning all labels in the subset are equal).\n", "\n", "For features with multiple categories values, splits are typically tested against the various one-vs-rest combinations, and for features with continuous values, these values are typically binned into ranges for testing.\n", "\n", "## Overfitting\n", "\n", "Trees resulting from the algorithm above are prone to becoming complex and specific to the training obversations, which means they will typically not generalise well. This is called overfitting - where the resulting tree (or model) is over-fitted to our training data.\n", "\n", "We can detect overfitting by comparing the performance of the model on the obversations it was constructed on (training data) to the performance on unmodelled observations (testing data). When the training error is less than the testing error, we know that the model is too specific to the training data and did not generalise well to the test data - and so it is overfitted.\n", "\n", "Overfitting can be offset by reducing the complexity of a tree, by establishing certain thresholds around the tree depth (or number of splits) and requirements for a split such as the minimum number of resulting observations on either side.\n", "\n", "## Cross Validation\n", "\n", "Like with any model, we want to build the best decision tree possible - that is, one which generalises well, and not overfitted as described above. If we can make more data available for training, the algorithms (by splitting on a measure such as *Gini impurity*) can possibly induce a better model. As we discovered previously, it's important to keep some amount (say, 20%) of labelled observations separate from the training process, so that we can produce a test score to determine how well the model performs on unseen data. As above, we can detect overfitting when the training error is less than the testing error.\n", "\n", "Well, it turns out that we can repeat our 80%/20% training/testing split more than once. For each repeat split, we simply take a different 20% of the data for testing. In fact, by repeating this process 5 times (5-fold) on different 20% tests, we can create 5 models and collectively use the *entire* dataset for training - while still providing a score of how well the model performs on unseen data.\n", "\n", "This general process is called *k-fold cross validation*, and in our case $k = 5$.\n", "\n", "\n", "## Implementing a Decision Tree" ] }, { "cell_type": "code", "execution_count": 83, "metadata": { "collapsed": false }, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjgAAAGXCAYAAAC+1ypQAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\niIiISMlRwBEREZGS8y8rjweBsFouRQAAAABJRU5ErkJggg==\n", "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "from sklearn.datasets import make_classification\n", "from sklearn.model_selection import cross_val_score\n", "from sklearn.model_selection import validation_curve\n", "from sklearn.tree import DecisionTreeClassifier\n", "import numpy as np\n", "\n", "# create some data and configure a decision tree\n", "X, y = make_classification(n_samples=1000, n_classes=2, n_features=20, n_informative=2)\n", "dt = DecisionTreeClassifier(criterion=\"gini\")\n", "\n", "# compute training and validation scores using 5-fold cv, for increasing values of max_depth\n", "max_depth_range = range(1,11)\n", "train_scores, valid_scores = validation_curve(dt, X, y, \"max_depth\", max_depth_range, scoring=\"roc_auc\", cv=5)\n", "\n", "# plot it\n", "%matplotlib inline\n", "import matplotlib.pyplot as plt\n", "\n", "plt.xkcd()\n", "plt.plot(max_depth_range, np.mean(train_scores,axis=1), label='training')\n", "plt.plot(max_depth_range, np.mean(valid_scores,axis=1), label='validation')\n", "plt.axhline(y=np.amax(np.mean(valid_scores,axis=1)), color='black', linestyle='--')\n", "plt.title('Decision Tree overfitting')\n", "plt.xlabel('max_depth')\n", "plt.ylabel('ROC AUC')\n", "plt.legend(loc=\"lower left\");" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Feature Reduction\n", "\n", "Decision trees can output the importance of features. This is a relative importance based on that particular feature's contribution to the overall tree, and is calculated by weighting the total proportion of observations it impacted and the resulting Gini impurities described above. While we won't go in depth on that calculation here, we can use the results to identify the key decision drivers of our model, and also identify the features which are not very important." ] }, { "cell_type": "code", "execution_count": 103, "metadata": { "collapsed": false }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[ 0.02216831 0.01659174 0.01985211 0.78566919 0.05045707 0.00782169\n", " 0.02084568 0.03703256 0.01629217 0.02326947]\n" ] }, { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiwAAAGXCAYAAACQkgvCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n", "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "from sklearn.datasets import make_classification\n", "from sklearn.model_selection import cross_val_score\n", "from sklearn.tree import DecisionTreeClassifier, export_graphviz\n", "\n", "# create some data and run decision tree with 5-fold CV\n", "X, y = make_classification(n_samples=1000, n_classes=2, n_features=10, n_informative=2)\n", "dt = DecisionTreeClassifier(criterion=\"gini\")\n", "dt.fit(X, y)\n", "print(dt.feature_importances_)\n", "\n", "# plot it\n", "%matplotlib inline\n", "import matplotlib.pyplot as plt\n", "\n", "plt.xkcd()\n", "plt.title('Feature Importance')\n", "plt.xlabel('Feature')\n", "plt.ylabel('Relative Importance')\n", "plt.bar(range(len(dt.feature_importances_)), dt.feature_importances_);" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "It's clear (usually\\*) from the bar chart, which features are and are not important. \\* Occasionally the noise introduced by `make_classification` in combination with overfitting in action can heighten the importance of uninformative features.\n", "\n", "## Performance with increasing observations\n", "\n", "In the last chatper, we noted that kNN performance will decline when the number of observations increases, and that we want to improve this with our decision tree model. To test this, we'll re-run the kNN test, alongside the same test for our decision tree." ] }, { "cell_type": "code", "execution_count": 115, "metadata": { "collapsed": false }, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAksAAAGXCAYAAACjqRTLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n", "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "from sklearn.datasets import make_classification\n", "from sklearn.model_selection import train_test_split\n", "from sklearn.neighbors import KNeighborsClassifier\n", "from sklearn.tree import DecisionTreeClassifier\n", "from sklearn import metrics\n", "import numpy as np\n", "\n", "px = []\n", "py1 = []\n", "py2 = []\n", "\n", "# this might take a while\n", "for samples in range(100,1000,200):\n", " \n", " # create some data and split into training and test sets\n", " X, y = make_classification(n_samples=samples, n_classes=2, n_features=50, n_informative=25)\n", " ts = 40/samples\n", " X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=ts, random_state=42)\n", "\n", " # fit kNN classifier and make predictions\n", " knn = KNeighborsClassifier(n_neighbors=4)\n", " knn.fit(X_train, y_train)\n", " \n", " # fit decision tree classifier and make predictions\n", " dt = DecisionTreeClassifier(criterion=\"gini\")\n", " dt.fit(X_train, y_train)\n", " \n", " # this is some Jupyter magic that allows us to capture execution timing\n", " # this might take a while to run\n", " knn_result = %timeit -o -q knn.predict_proba(X_test)\n", " dt_result = %timeit -o -q dt.predict_proba(X_test)\n", " \n", " px.append(samples)\n", " py1.append(knn_result.best)\n", " py2.append(dt_result.best)\n", "\n", "%matplotlib inline\n", "import matplotlib.pyplot as plt\n", "\n", "plt.xkcd()\n", "plt.plot(px,py1,label='knn')\n", "plt.plot(px,py2,label='decision tree')\n", "plt.title('kNN vs Decision tree scoring performance')\n", "plt.xlabel('Number of training samples')\n", "plt.ylabel('Execution time')\n", "plt.legend(loc=\"upper left\");" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "As you can see, scoring time for the decision tree remains uniform and insignificant compared to the kNN classifier.\n", "\n", "## What next?\n", "\n", "So now we've looked at an algorithm that is able to both consider complex relationships between features, but also create a model on the training step that is powerful enough to reduce scoring time, but can be prone to overfitting without careful attention being given to (hyper) parameters.\n", "\n", "One alternative which is less prone to overfitting is the Logistics Regression classifier, which instead of attempting to model complex decision rules within a dataset, instead opts to find the simplest linear boundary between classes." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "---\n", "\n", "[Table of Contents](00.00-Learning-ML.ipynb#Table-of-Contents) • [← *Chapter 2.03 - k-Nearest Neighbours*](02.03-k-Nearest-Neighbours.ipynb) • [*Chapter 2.05 - Logistic Regression* →](02.05-Logistic-Regression.ipynb)\n", "\n", "" ] } ], "metadata": { "anaconda-cloud": {}, "kernelspec": { "display_name": "Python [default]", "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": "3.5.2" } }, "nbformat": 4, "nbformat_minor": 1 }