{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Toy Example: Ridge Regression vs. SVM\n",
"
\n",
" \n",
"
\n",
" In this toy example we will compare two machine learning models: Ridge Regression and C-SVM. The data is generated in silico and is only used to illustrate how to use Ridge Regression and C-SVM.\n",
"
\n",
"\n",
"## Problem Description of the Toy Example\n",
"\n",
" \n",
"
\n",
"A new cancer drug was developed for therapy. During the clinical trail the researchers releaized that the drug had a faster response for a certain subgroup of the patients, while it was less responsive in the others. In addition, the researchers recognized that the drug leads to severe side-effects the longer the patient is treated with the drug. The goal should be to reduce the side effects by treating only those patients that are predicted to have a fast response when taking the drug.\n",
"
\n",
" \n",
"
\n",
"The researches believe that different genetic mutations in the genomes of the individual patients might play a role for the differences in response times.\n",
"
\n",
" \n",
"
\n",
" The researches contacted the machine learning lab to build a predictive model. The model should predict the individual response time of the drug based on the individual genetic backgrounds of a patient.\n",
"
\n",
" \n",
"
\n",
"For this purpose, we get a dataset of 400 patients. For each patient a panel of 600 genetic mutations was measured. In addition, the researchers measured how many days it took until the drug showed a positive response.\n",
"
\n",
"\n",
"\n",
"## 1. Using Ridge Regression to predict the response time\n",
"
\n",
" To predict the response time of the drug for new patients, we will train a Ridge Regression model. The target variable for this task is the response time in days. The features are the 600 genetic mutations measured for each of the 400 patients. To avoid overfitting we will use a nested-crossvalidation to determine the optimal hyperparamter.\n",
"
\n",
"### 1.1 Data Preprocessing"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"scrolled": true
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Orginal Data\n",
"Number Patients:\t400\n",
"Number Features:\t600\n",
"\n",
"Training Data\n",
"Number Patients:\t320\n",
"Number Features:\t600\n",
"\n",
"Testing Data\n",
"Number Patients:\t80\n",
"Number Features:\t600\n"
]
}
],
"source": [
"%matplotlib inline\n",
"import scipy as sp\n",
"import matplotlib\n",
"import pylab as pl\n",
"matplotlib.rcParams.update({'font.size': 15})\n",
"\n",
"from sklearn.linear_model import Ridge\n",
"from sklearn.svm import SVC\n",
"from sklearn.model_selection import KFold, StratifiedKFold, GridSearchCV,StratifiedShuffleSplit\n",
"from sklearn.model_selection import cross_val_score, train_test_split\n",
"from sklearn.metrics import accuracy_score, mean_squared_error, mean_absolute_error\n",
"from sklearn.metrics import roc_curve, auc\n",
"\n",
"def visualized_variance_bias_tradeoff(hyperp, line_search, optimal_hyperp,classification=False):\n",
" pl.figure(figsize=(18,7))\n",
" if classification:\n",
" factor=1\n",
" else:\n",
" factor=-1\n",
" pl.plot(hyperp,line_search.cv_results_['mean_train_score']*factor,label=\"Training Error\",color=\"#e67e22\")\n",
" pl.fill_between(hyperp,\n",
" line_search.cv_results_['mean_train_score']*factor-line_search.cv_results_['std_train_score'],\n",
" line_search.cv_results_['mean_train_score']*factor+line_search.cv_results_['std_train_score'],\n",
" alpha=0.3,color=\"#e67e22\")\n",
" pl.plot(hyperp,line_search.cv_results_['mean_test_score']*factor,label=\"Validation Error\",color=\"#2980b9\")\n",
" pl.fill_between(hyperp,\n",
" line_search.cv_results_['mean_test_score']*factor-line_search.cv_results_['std_test_score'],\n",
" line_search.cv_results_['mean_test_score']*factor+line_search.cv_results_['std_test_score'],\n",
" alpha=0.3,color=\"#2980b9\")\n",
" pl.xscale(\"log\")\n",
" if classification:\n",
" pl.ylabel(\"Accuracy\")\n",
" else:\n",
" pl.ylabel(\"Mean Squared Error\")\n",
" pl.xlabel(\"Hyperparameter\")\n",
" pl.legend(frameon=True)\n",
" pl.grid(True)\n",
" pl.axvline(x=optimal_hyperp,color='r',linestyle=\"--\")\n",
" pl.title(\"Training- vs. Validation-Error (Optimal Hyperparameter = %.1e)\"%optimal_hyperp);\n",
"\n",
"random_state = 42\n",
"\n",
"#Load Data\n",
"data = sp.loadtxt(\"data/X.txt\")\n",
"binary_target = sp.loadtxt(\"data/y_binary.txt\")\n",
"continuous_target = sp.loadtxt(\"data/y.txt\")\n",
"\n",
"#Summary of the Data\n",
"print(\"Orginal Data\")\n",
"print(\"Number Patients:\\t%d\"%data.shape[0])\n",
"print(\"Number Features:\\t%d\"%data.shape[1])\n",
"print()\n",
"\n",
"#Split Data into Training and Testing data\n",
"train_test_data = train_test_split(data,\n",
" continuous_target,\n",
" test_size=0.2,\n",
" random_state=random_state)\n",
"training_data = train_test_data[0]\n",
"testing_data = train_test_data[1]\n",
"training_target = train_test_data[2]\n",
"testing_target = train_test_data[3]\n",
"\n",
"print(\"Training Data\")\n",
"print(\"Number Patients:\\t%d\"%training_data.shape[0])\n",
"print(\"Number Features:\\t%d\"%training_data.shape[1])\n",
"print()\n",
"print(\"Testing Data\")\n",
"print(\"Number Patients:\\t%d\"%testing_data.shape[0])\n",
"print(\"Number Features:\\t%d\"%testing_data.shape[1])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 1.2 Train Ridge Regression on training data\n",
"\n",
"The first step is to train the ridge regression model on the training data with a **5-fold cross-validation** with an **internal line-search** to find the **optimal hyperparameter $\\alpha$**. We will plot the **training errors** against the **validation errors**, to illustrate the effect of different $\\alpha$ values."
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"5-fold nested cross-validation\n",
"Mean-Squared-Error:\t\t587.09 (-+ 53.45)\n",
"\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"