{ "cells": [ { "cell_type": "markdown", "metadata": { "collapsed": true }, "source": [ "This notebook demonstrates how to use the `RBFScanner` object in `scanners.py` for interactive optimization of support vector classifier (SVC) fitting in scikit-learn. Currently, this approach is limited to optimizing the C and gamma parameters for a Gaussian (or \"radial basis function\", RBF) kernel, which allows for non-linear decision boundary functions and functions well in the vast majority of cases.\n", "\n", "The non-Python details of this process, including a description of the parameters and demonstration of their are consequences, are discussed [here](http://thomasbkahn.github.io/svc-opt).\n", "\n", "The mechanics of SVC operation are beyond the scope of this demonstration, but they are described excellently in this [tutorial by Jake VanderPlas](https://www.youtube.com/watch?v=L7R4HUQ-eQ0) (from 1:45:50 to 2:05:30)." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Loading Sample Data\n", "\n", "This demonstration will utilize the [famous Iris dataset](https://en.wikipedia.org/wiki/Iris_flower_data_set), which is commonly used in data science and machine learning examples. This dataset consists of measurements of four anatomical features from 50 examples each of three different species of Iris flower. The classification task is to utilize these four measurements to determine which of the three species a sample belongs to.\n", "\n", "The dataset is available through scikit-learn (and many other sources).\n", "\n", "The Iris data will be loaded into a pandas `DataFrame` object to make use of seaborn plotting functionality, and each of the four features will be scaled independently by computing the Z score. The `DataFrame` is not a required intermediate for SVC optimization, but feature scaling is a *very important* preprocessing step." ] }, { "cell_type": "code", "execution_count": 1, "metadata": { "collapsed": true }, "outputs": [], "source": [ "# imports\n", "import pandas as pd\n", "from scipy import stats\n", "from sklearn.datasets import load_iris" ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "collapsed": false }, "outputs": [], "source": [ "# load and make Iris DataFrame\n", "iris_data = load_iris()\n", "iris_df = pd.DataFrame(iris_data['data'], columns=iris_data['feature_names'])\n", "iris_data['target_names'] = [name[0].upper()+name[1:] for name in iris_data['target_names']]\n", "iris_df['species'] = [iris_data['target_names'][target_i] for target_i in iris_data['target']]\n", "\n", "# prepare normalized DataFrame\n", "feature_cols = [col for col in iris_df.columns if col.endswith('(cm)')]\n", "norm_df = iris_df.copy()\n", "norm_df[feature_cols] = iris_df[feature_cols].apply(stats.zscore, axis=0)\n", "norm_df.columns = [col.replace('(cm)', 'normalized') for col in norm_df.columns]\n", "\n", "# extract data and other necessities\n", "feature_cols = [col for col in norm_df.columns if col.endswith('normalized')]\n", "iris_X = norm_df[feature_cols].values\n", "iris_y = iris_data['target']\n", "iris_class_names = iris_data['target_names']" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Data Visualization" ] }, { "cell_type": "code", "execution_count": 3, "metadata": { "collapsed": true }, "outputs": [], "source": [ "# imports\n", "import warnings\n", "# to prevent harmless warnings raised by matplotlib\n", "warn_msg = (\n", " \"axes.color_cycle is deprecated and replaced \"\n", " \"with axes.prop_cycle; please use the latter.\"\n", ")\n", "warnings.filterwarnings(\"ignore\", message=warn_msg)\n", "import seaborn as sns\n", "%matplotlib inline\n", "sns.set_style({'axes.facecolor':'wheat'})" ] }, { "cell_type": "code", "execution_count": 4, "metadata": { "collapsed": false }, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAyAAAALJCAYAAABMTEeWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n", "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "sns.pairplot(norm_df, hue='species', diag_kind='kde');" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "As seen above, Setosa can easily be separated (even with just the petal-related features), but the separation between Versicolor and Virginica will be a bit trickier." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# `RBFScanner` Objects\n", "\n", "Ranges of gamma and C values will be scanned to determine the optimal parameter settings. To accomplish this, the `RBFScanner` object is used, which scans through the parameter values, fits an RBF-kernel support vector classifier using a 0.67/0.33 random train/test splitting of the data (repeated for a specified number of iterations), and stores the weighted-average values of various performance metrics in arrays.\n", "\n", "The only required arguments for `RBFScanner` are `X` (input data with n_samples by n_features shape) and `y` (the input labels).\n", "\n", "The optional arguments (with their default values) are as follows:\n", "1. `C_lims=(-12,12)`. The range of C values to scan.\n", "2. `gamma_lims=(-12,12)`. The range of gamma values for to scan.\n", "3. `n_steps=50`. The number of steps in the C and gamma ranges.\n", "4. `n_iters=20`. The number of repeats of the train/test splitting and classifier fitting process. Higher values decrease the noise of the results, but cause the scans to take longer.\n", "5. `logvals=True`. Whether the the values of `C_lims` and `gamma_lims` are on a logarithmic scale.\n", "6. `class_names=None`. A list of the names of each class. If not provided, the classes will be referred to as \"Class 0\", \"Class 1\", etc.\n", "7. `seed`. Random seed to generate list of 1000 random seeds, each of which will be used to generate list of `n_iters` $\\times$ `n_steps` $\\times$ `n_steps` seeds for train/test split. This is useful for reproducing results." ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "collapsed": false }, "outputs": [], "source": [ "# import\n", "from scanners import RBFScanner" ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "collapsed": false }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [] } ], "source": [ "# create object (scan will run upon initialization)\n", "# Warning: this will take several minutes for n_iters=60,\n", "# use lower value for faster scans.\n", "iris_scan = RBFScanner(iris_X, iris_y, class_names=iris_class_names, n_iters=60, seed=12)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Visualizing Results\n", "\n", "## Built-in Plotting Methods\n", "\n", "To facilitate interactive optimization of parameters, built-in Bokeh plotting methods are provided. These methods are:\n", "\n", "- `show_single()`, which shows a plot of a single performance metric.\n", "- `show_train_test()`, which shows the respective accuracies of the training and test data, as well as their difference. This is useful for determining where the classifiers overfit or underfit the data.\n", "- `show_classes()`, which shows the overall performance (*i.e.* accuracy of classification upon the test set), as well as the performance among individual classes (given by the [$F_1$](https://en.wikipedia.org/wiki/F1_score) score for that class).\n", "\n", "The three plotting methods are called in a similiar manner, though `show_single()` takes a required `title` argument, which is a string that indicates which of the result matrices is to be demonstrated. Possible options for this value can be found in the `titles` attribute. Additionally, the number of columns to place the plots in the multiple plot methods can be specified with the optional `n_cols` parameter.\n", "\n", "All of the plotting methods also have the following additional optional parameters:\n", "1. `dims` : Dimension of the individual Bokeh plot objects (as a (width, height) tuple) in units of pixels.\n", "2. `v_lims` : Limits for the color scale (default is 0.0 to 1.0).\n", "3. `save_string` : If left as default (`None`), the plot will output to the ipython notebook that it was called from. If `return`, the grouped `gridplot` object will be returned (which is useful for inline embedding). If any string that ends with `.html`, the figure will be saved to a file with the corresponding filename." ] }, { "cell_type": "code", "execution_count": 7, "metadata": { "collapsed": false }, "outputs": [ { "data": { "text/plain": [ "['Overall Accuracy',\n", " 'Training Accuracy',\n", " 'Difference',\n", " 'Setosa (F1 Score)',\n", " 'Versicolor (F1 Score)',\n", " 'Virginica (F1 Score)']" ] }, "execution_count": 7, "metadata": {}, "output_type": "execute_result" } ], "source": [ "iris_scan.titles" ] }, { "cell_type": "code", "execution_count": 8, "metadata": { "collapsed": false }, "outputs": [ { "data": { "text/html": [ "\n", "\n", " \n", "
\n", " \n", " BokehJS successfully loaded.\n", "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "\n", "\n", "
\n", "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# try replacing with any value in iris_scan.titles\n", "title = iris_scan.titles[0]\n", "\n", "iris_scan.show_single(title)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "By default, values that are identical to the maximum (or are within a set tolerance, given by the `tol` attribute with a default value of 0.001) are highlighted in the plot, and the score result in the tooltip is shown in a distinct color. These colors can be changed by inputting any web color or hexadecimal RGB string to the `plot_params` attribute dictionary, or turned off by entering `None` for either value. The colormap can be changed using the `set_cmap` method, which takes a matplotlib colormap object as input. Defaults can be restored by calling the `restore_plot_defaults` method." ] }, { "cell_type": "code", "execution_count": 9, "metadata": { "collapsed": false }, "outputs": [ { "data": { "text/html": [ "\n", "\n", " \n", "
\n", " \n", " BokehJS successfully loaded.\n", "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "\n", "\n", "
\n", "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "import matplotlib.pyplot as plt\n", "\n", "# set colormap to viridis\n", "iris_scan.set_cmap(plt.cm.viridis)\n", "\n", "# set highlight color to firebrick\n", "iris_scan.plot_params['highlight_max'] = 'firebrick'\n", "\n", "# set tooltip highlighted text to forestgreen\n", "iris_scan.plot_params['highlight_hovertext'] = 'forestgreen'\n", "\n", "iris_scan.show_single(title)" ] }, { "cell_type": "code", "execution_count": 10, "metadata": { "collapsed": false }, "outputs": [ { "data": { "text/html": [ "\n", "\n", " \n", "
\n", " \n", " BokehJS successfully loaded.\n", "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "\n", "\n", "
\n", "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# restore defaults\n", "iris_scan.restore_plot_defaults()\n", "\n", "# turn off both plot and text highlighting\n", "iris_scan.plot_params['highlight_hovertext'] = None\n", "iris_scan.plot_params['highlight_max'] = None\n", "\n", "# using narrower v_lims to better visualize structure\n", "iris_scan.show_single(title, v_lims=(0.9, 1.0))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Overfitting versus Underfitting" ] }, { "cell_type": "code", "execution_count": 11, "metadata": { "collapsed": false }, "outputs": [ { "data": { "text/html": [ "\n", "\n", " \n", "
\n", " \n", " BokehJS successfully loaded.\n", "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "\n", "\n", "
\n", "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "iris_scan.restore_plot_defaults()\n", "iris_scan.plot_params['highlight_max'] = None\n", "iris_scan.show_train_test()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The training accuracy is 1.0 for most cases where gamma is $>0.1$ and C is $>1.0$. However, the overall (test) accuracy is near the baseline value in most of this region, causing the difference signal to be large. This corresponds to overfitting, since the training set is fit perfectly with a model that generalizes poorly to new examples. In the area in which the test accuracy is high, the training accuracy is also correspondingly high (and in fact, slightly higher, as all models will in general perform better on the training set than the test set) leading to a low difference signal. The remaining region of the plots correspond to underfitting, as the resulting classifier models perform poorly for both the training and test sets of data." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Performance for Individual Classes" ] }, { "cell_type": "code", "execution_count": 12, "metadata": { "collapsed": false }, "outputs": [ { "data": { "text/html": [ "\n", "\n", " \n", "
\n", " \n", " BokehJS successfully loaded.\n", "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "\n", "\n", "
\n", "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "iris_scan.restore_plot_defaults()\n", "iris_scan.show_classes()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The overall performance is largely correlated with the individual performance for Virginica or Versicolor classification, as these two classes were shown to be the most convoluted of the three. Setosa classification is often perfect, which is unsurprising given the distance between that class and the others in the feature-space, as shown above." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Iteratively Narrowing Parameter Range\n", "\n", "To determine more-robust and precise guidelines on the values for C and gamma, the scan can be iterated over an increasingly narrow range of values, increasing the effective resolution. At each step, the performance over the range is evaluated, and the region that reasonably appears to contain a suitable optimum is chosen to determine new limits for the next, smaller scan. (Given the degree of noise in the performance metrics, a more robust method to select the optimum could incorporate polynomial regression on the output metric as a function of input C and gamma values.)\n", "\n", "From the above results, it's clear that the classifier performs well almost exclusively in the upper left quadrant. Therefore, the next scan will cover only this region." ] }, { "cell_type": "code", "execution_count": 13, "metadata": { "collapsed": false }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [] } ], "source": [ "# rescan with smaller range\n", "iris_scan.scan(C_lims=(0, 12), gamma_lims=(-12, 0))" ] }, { "cell_type": "code", "execution_count": 14, "metadata": { "collapsed": false, "scrolled": false }, "outputs": [ { "data": { "text/html": [ "\n", "\n", " \n", "
\n", " \n", " BokehJS successfully loaded.\n", "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "\n", "\n", "
\n", "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# visualize per-class performance, using narrower\n", "# range on v_lims to better illustrate structure\n", "# near maximal values\n", "iris_scan.show_classes(v_lims=(0.9, 1.0))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The best performing classifiers (in terms of combined overall, Versicolor, and Virginica performances) on average were those with a value of $7.197\\times10^{-6}$ for gamma and $4.498\\times10^{4}$ for C, with an average overall accuracy of 97.4 % (up from 97.2 %, in the previous scan). The next scan will consider 2 orders of magnitude for each parameter, centered around this point." ] }, { "cell_type": "code", "execution_count": 15, "metadata": { "collapsed": false }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [] } ], "source": [ "iris_scan.scan(C_lims=(3, 5), gamma_lims=(-6, -4))" ] }, { "cell_type": "code", "execution_count": 16, "metadata": { "collapsed": false }, "outputs": [ { "data": { "text/html": [ "\n", "\n", " \n", "
\n", " \n", " BokehJS successfully loaded.\n", "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "\n", "\n", "
\n", "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "iris_scan.show_classes(v_lims=(0.9, 1.0))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The best performing parameters were $1.526\\times10^{-5}$ (gamma) and $2.442\\times10^{4}$ (C). For the final scan, linear ranges around these values will be used." ] }, { "cell_type": "code", "execution_count": 17, "metadata": { "collapsed": false }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [] } ], "source": [ "iris_scan.scan(C_lims=(2e4, 3e4), gamma_lims=(1e-5, 2e-5), logvals=False)" ] }, { "cell_type": "code", "execution_count": 18, "metadata": { "collapsed": false }, "outputs": [ { "data": { "text/html": [ "\n", "\n", " \n", "
\n", " \n", " BokehJS successfully loaded.\n", "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "\n", "\n", "
\n", "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "iris_scan.show_classes(v_lims=(0.9, 0.98))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The maximum overall accuracy has improved to 97.5 %, and can be reached through several different pairs of C and gamma values within this range. One of these pairs (gamma = $1.245\\times10^{-5}$, C = $2.347\\times10^{4}$) also maximizes performance in classifying Versicolor and Virginica individually, and thus represents optimal paramater values for future classifier training." ] } ], "metadata": { "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": "3.5.1" } }, "nbformat": 4, "nbformat_minor": 0 }