{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Machine Learning Example\n", "\n", "### Authors: \n", "- Christian Michelsen (Niels Bohr Institute)\n", "- Troels C. Petersen (Niels Bohr Institute)\n", "\n", "### Date: \n", "- 16-11-2018 (latest update)\n", "\n", "***\n", "\n", "Python macro for analysing more complex challenge of classifying events (in 2D) using Machine Learning.\n", "\n", "***\n", "\n", "First, we import the modules we want to use:" ] }, { "cell_type": "code", "execution_count": 18, "metadata": {}, "outputs": [], "source": [ "import numpy as np\n", "import matplotlib.pyplot as plt\n", "from ipywidgets import interactive # to make plots interactive" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "And set the parameters of the notebook:" ] }, { "cell_type": "code", "execution_count": 19, "metadata": {}, "outputs": [], "source": [ "r = np.random\n", "r.seed(42)\n", "\n", "export_tree = True\n", "plot_fisher_discriminant = True\n", "\n", "test_point = np.array([0, 0.5]).reshape(1, -1)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Functions:\n", "\n", "First we define `get_corr_pars` which generate (linearly) correlated numbers:" ] }, { "cell_type": "code", "execution_count": 20, "metadata": { "lines_to_next_cell": 2 }, "outputs": [], "source": [ "def get_corr_pars(mu1, mu2, sig1, sig2, rho12) : \n", "\n", " # Transform to correlated random numbers (see Barlow page 42-44):\n", " # This is nothing more than a matrix multiplication written out!!!\n", " # Note that the absolute value is taken before the square root to avoid sqrt(x) with x<0.\n", " theta = 0.5 * np.arctan( 2.0 * rho12 * sig1 * sig2 / ( sig1**2 - sig2**2 ) )\n", " sigu = np.sqrt( np.abs( ((sig1*np.cos(theta))**2 - (sig2*np.sin(theta))**2 ) / ( np.cos(theta)**2 - np.sin(theta)**2) ) )\n", " sigv = np.sqrt( np.abs( ((sig2*np.cos(theta))**2 - (sig1*np.sin(theta))**2 ) / ( np.cos(theta)**2 - np.sin(theta)**2) ) )\n", "\n", " u = r.normal(0.0, sigu)\n", " v = r.normal(0.0, sigv)\n", "\n", " # Transform into (possibly) correlated random Gaussian numbers x1 and x2:\n", " x1 = mu1 + np.cos(theta)*u - np.sin(theta)*v\n", " x2 = mu2 + np.sin(theta)*u + np.cos(theta)*v\n", "\n", " return x1, x2" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Define the function `plot_decision_regions` which plots decision boundaries:" ] }, { "cell_type": "code", "execution_count": 21, "metadata": { "lines_to_next_cell": 2 }, "outputs": [], "source": [ "from matplotlib.colors import ListedColormap\n", "from sklearn.metrics import roc_curve, auc\n", "\n", "def plot_decision_regions(X, y, classifier, resolution=0.02, title=None, fig=None, ax=None):\n", " \n", " # define colors\n", " colors = ('red', 'blue')\n", " cmap = ListedColormap(colors)\n", " \n", " # define signal and background\n", " sig = X[y == 1]\n", " bkg = X[y == 0]\n", " \n", " # compute the decision surface\n", " x1_min, x1_max = X[:, 0].min() - 1, X[:, 0].max() + 1\n", " x2_min, x2_max = X[:, 1].min() - 1, X[:, 1].max() + 1\n", " xx1, xx2 = np.meshgrid(np.arange(x1_min, x1_max, resolution), np.arange(x2_min, x2_max, resolution))\n", " Z = classifier.predict(np.array([xx1.ravel(), xx2.ravel()]).T)\n", " Z = Z.reshape(xx1.shape)\n", " \n", " # set up the figure\n", " if fig is None and ax is None:\n", " fig, ax = plt.subplots(1, 2, figsize=(16, 8))\n", " \n", " # plot the decision surface and plot individual points on ax[0]\n", " ax[0].contourf(xx1, xx2, Z, alpha=0.2, cmap=cmap)\n", " ax[0].scatter(sig[:, 0], sig[:, 1], s=4, c='blue', label='sig', alpha=0.3)\n", " ax[0].scatter(bkg[:, 0], bkg[:, 1], s=4, c='red', label='bkg', alpha=0.3)\n", "\n", " ax[0].set(xlim=(xx1.min(), xx1.max()), ylim=(xx2.min(), xx2.max()), xlabel='Parameter A', ylabel='Parameter B')\n", " \n", " \n", " # predict and plot the prediction of the test point on ax[0]\n", " z_test = classifier.predict(test_point)[0]\n", " if z_test == 0:\n", " color = 'red'\n", " else:\n", " color = 'blue'\n", " ax[0].scatter(test_point[0,0], test_point[0,1], c='w', s=200, marker='o')\n", " ax[0].scatter(test_point[0,0], test_point[0,1], c=color, s=150, marker='*')\n", " \n", " # set the legend on ax[0]\n", " ax[0].legend()\n", " \n", " \n", " # set up ax[1]:\n", " \n", " # compute y prediction probabilities:\n", " y_predicted_proba = classifier.predict_proba(X)[:, 1]\n", " \n", " # Compute ROC curve and ROC area\n", " FPR, TPR, _ = roc_curve(y, y_predicted_proba)\n", " roc_auc = auc(FPR, TPR)\n", " \n", " lw = 2\n", "\n", " # plot the ROC curve\n", " ax[1].plot(FPR, TPR, color='darkorange', lw=lw, label='ROC curve (area = %0.3f)' % roc_auc)\n", " ax[1].plot([0, 1], [0, 1], color='navy', lw=lw, linestyle='--')\n", " ax[1].set(xlim=[-0.01, 1.0], ylim=[-0.01, 1.05], xlabel='False Positive Rate', ylabel='True Positive Rate')\n", " ax[1].legend(loc=\"lower right\")\n", " \n", " if title:\n", " ax[0].set(title=title)\n", " ax[1].set(title=title)\n", " \n", " return fig, ax" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Below we define `animate_ML_estimator_generator` which takes an estimator, fits it given the specified keywords and plots the decision regions:" ] }, { "cell_type": "code", "execution_count": 22, "metadata": {}, "outputs": [], "source": [ "def animate_ML_estimator_generator(clf, title, X, y, **kwargs): \n", " estimator = clf(**kwargs)\n", " estimator = estimator.fit(X, y)\n", " plot_decision_regions(X, y, classifier=estimator, title=title.capitalize())" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Generate and plot data:\n", "\n", "Set number of data points to generate and parameters:" ] }, { "cell_type": "code", "execution_count": 23, "metadata": {}, "outputs": [], "source": [ "N = 10000\n", "\n", "mu_sig_A = 0\n", "mu_sig_B = 1\n", "sigma_sig_A = 1.5 \n", "sigma_sig_B = 2\n", "rho_sig = 0.8\n", "\n", "mu_bkg_A = -2\n", "mu_bkg_B = 4\n", "bkgma_bkg_A = 1.5 \n", "bkgma_bkg_B = 2\n", "rho_bkg = 0.8" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Generate correlated parameters:" ] }, { "cell_type": "code", "execution_count": 24, "metadata": {}, "outputs": [], "source": [ "sig = np.zeros((N, 2))\n", "for i in range(N):\n", " x1, x2 = get_corr_pars(mu_sig_A, mu_sig_B, sigma_sig_A, sigma_sig_B, rho_sig) \n", " sig[i, :] = x1, x2\n", "\n", "bkg = np.zeros((N, 2))\n", "for i in range(N):\n", " x1, x2 = get_corr_pars(mu_bkg_A, mu_bkg_B, bkgma_bkg_A, bkgma_bkg_B, rho_bkg) \n", " bkg[i, :] = x1, x2" ] }, { "cell_type": "markdown", "metadata": { "lines_to_next_cell": 2 }, "source": [ "Plot the generated random numbers:" ] }, { "cell_type": "code", "execution_count": 25, "metadata": { "lines_to_next_cell": 2 }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "Nbins = 100\n", "\n", "fig, ax = plt.subplots(1, 4, figsize=(18, 5))\n", "\n", "ax[0].hist(sig[:, 0], Nbins, histtype='step', label='sig', color='blue')\n", "ax[0].hist(bkg[:, 0], Nbins, histtype='step', label='bkg', color='red')\n", "ax[0].set(xlabel='Variable A', ylabel='Counts', title='Histogram of Variable A')\n", "ax[0].legend()\n", "\n", "ax[1].hist(sig[:, 1], Nbins, histtype='step', label='sig', color='blue')\n", "ax[1].hist(bkg[:, 1], Nbins, histtype='step', label='bkg', color='red')\n", "ax[1].set(xlabel='Variable B', ylabel='Counts', title='Histogram of Variable B')\n", "ax[1].legend()\n", "\n", "ax[2].scatter(sig[:, 0], sig[:, 1], s=4, c='blue', label='sig', alpha=0.5)\n", "ax[2].scatter(bkg[:, 0], bkg[:, 1], s=4, c='red', label='bkg', alpha=0.5)\n", "ax[2].set(xlabel='Variable A', ylabel='Variable B', title='Scatterplot of Variable A and B')\n", "ax[2].legend();" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Fishers Discriminant: \n", "How to perform a __[Fishers Linear Discimant Analysis](https://scikit-learn.org/stable/modules/generated/sklearn.discriminant_analysis.LinearDiscriminantAnalysis.html)__ in Scikit Learn. \n", "\n", "\n", "\n", "First load the proper package:" ] }, { "cell_type": "code", "execution_count": 26, "metadata": {}, "outputs": [], "source": [ "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis as LDA" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Then we convert the two numpy arrays one big one called `X`. `y` denotes the class (1 for signal, 0 for background)" ] }, { "cell_type": "code", "execution_count": 27, "metadata": {}, "outputs": [], "source": [ "X = np.vstack((sig, bkg))\n", "\n", "y = np.zeros(2*N) \n", "y[:N] = 1" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now do the Linear Discriminant Analysis (LDA):" ] }, { "cell_type": "code", "execution_count": 28, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "LDA coefficients: [[ 1.03939976]\n", " [-0.79035145]]\n" ] } ], "source": [ "# initialise the LDA method\n", "sklearn_lda = LDA(n_components=2)\n", "\n", "# fit the data\n", "sklearn_lda.fit(X, y)\n", "\n", "# transform the data\n", "X_lda_sklearn = sklearn_lda.transform(X) \n", "\n", "print(f\"LDA coefficients: {sklearn_lda.scalings_}\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Extract the tranformed variables" ] }, { "cell_type": "code", "execution_count": 29, "metadata": {}, "outputs": [], "source": [ "sig_lda = X_lda_sklearn[y == 1]\n", "bkg_lda = X_lda_sklearn[y == 0]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "and plot it on the figure:" ] }, { "cell_type": "code", "execution_count": 30, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "execution_count": 30, "metadata": {}, "output_type": "execute_result" } ], "source": [ "ax[3].hist(sig_lda, Nbins, histtype='step', label='sig', color='red')\n", "ax[3].hist(bkg_lda, Nbins, histtype='step', label='bkg', color='blue')\n", "ax[3].set(xlabel='Fisher Discriminant', ylabel='Counts', title='Fishers discriminant')\n", "ax[3].legend()\n", "\n", "fig.tight_layout()\n", "\n", "fig" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# What if the dataset was completely different?\n", "\n", "Now we want to eksamine the dataset given in `DataSet_ML.txt`. First we load it, extract the relevant data and plot it:\n" ] }, { "cell_type": "code", "execution_count": 35, "metadata": {}, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAABJQAAAHwCAYAAADw9zWuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzs3Xu8ZGdZJ/rfQzoYLuES6MnEdNgVleuABGwiCgeRiyQoBGaQIXKJDtKOgqOj3MejCDNzdA4Dcs4gThAlzshN0CFgABEDCMOtuYcEDhH3lg6BNJGIEYJGn/NHVSe7q3d37+rsqtq19/f7+axP13rX7alVq/db9ax3vW91dwAAAABgvW427wAAAAAAWCwSSgAAAABMREIJAAAAgIlIKAEAAAAwEQklAAAAACYioQQAAADARCSUOGZV9ZmqevC845inqnpsVX2xqq6tqvvM4Hj/R1V9bp3r/nhVve8Iy99dVT+5cdEBHEw9oZ4AALYuCSXWVFXLVfWwsbKDvnh297/o7ncfZT+Dquqq2jGlUOftxUme0d237u6Pr15QVZ+tqn8zvkFV/VxV7T2Wg3X3n3f3XY8x1g1VVbce/UB627xjAWZPPbFu266eWPWZXjuavlJVv1lVx88zLgBgY0kosdA2wQ+QpSSfOcyyC5I8ZY3yJ4+WTWQTvNdx/yrJt5I8vKr++byDAVjLJvjbuZ3ridt1962T3CvJ9yV5+pzjAQA2kIQSx2z13emqOrOq9lbV10d3Il8yWu29o3+vGd2l/L6qullV/VJVrVTVVVX1e1V121X7fcpo2dVV9X+OHecFVfXGqvqfVfX1JD8+OvYHquqaqrqyqv5bVd181f66qn6mqj5fVX9bVS+qqu+sqv89ivcNq9cfe49rxlpV31ZV1yY5Lsknq+ov1tj8fyR5YFUtrdrfPZJ8d5LXjuZ/oqouG8X1har6qVXrPriq9lXVc6rqy0l+90DZqnWeW1V/Mdr+0qp67KFvof5bVf3N6E74Q4/wef6bUSxfq6p3rI77MM5L8ltJPpXkSUdZF9iG1BPbvp5IknT3VUnemeQe61kfAFgMEkpslJcleVl33ybJdyZ5w6j8QaN/bzdq7v+BJD8+mn4wyXckuXWS/5bc8EX6N5M8MckpSW6b5NSxY52T5I1Jbpfk95P8Y5J/n+SOGd4BfWiSnxnb5hFJvifJ/ZM8O8n5GSZBTktyzyTnHuZ9rRlrd39rdNc1Se7d3d85vmF370tycYZ3mg94cpKLuvuro/mrkvxIktsk+YkkL62q+65a/58nOSnDO9x71ojvL5L8Hxmep19N8j+r6pRVy793tM4dk/xKkj+sqpPGd1JV5yR5fpJ/mWRnkj/P6MfMWkY/Ih6c4fn//ax9hx1gNfXEmK1cT4xt++0Znt8Prmd9AGAxSChxJP9rdDf3mqq6JsMv8IfzD0m+q6ru2N3XdveRvjQ+MclLuvsL3X1tkucleUINm+o/Lslbuvt93f33SX45SY9t/4Hu/l/d/U/d/c3u/mh3f7C7r+/u5ST/PckPjG3zX7r76939mSSXJPmT0fH/Jsnbkhyuo9QjxboeF2T0Q6Gqbjba3w2PMXT3H3f3X/TQe5L8SYZf/A/4pyS/Mvph8s3xnXf3H3T3l0bn4vVJPp/kzFWrXJXkN7r7H0bLP5fkh9eI898m+b+6+7Luvj7Jf05yxhHuPj85yae6+9Ikr0vyL2oGnc0Cm456Qj1xpFZKXx1dF1ck+bsMk3wAwBYhocSRPKa7b3dgyqF3c1d7apK7JPlsVX2kqn7kCOt+e5KVVfMrSXYkOXm07IsHFnT3N5JcPbb9F1fPVNVdquqtVfXl0eMN/znDO62rfWXV62+uMX/rrO1Isa7HHyY5parun2GLnlsm+eNVsZ9dVR+sqr8efel+5Fjs+7v7usPtfPTYxydW/Zi759j2V3T36h9aK6P3NG4pyctW7eevk1QOvet/wFMyvOuf7r4iyXsyfAQO2F7UE+qJw9UTSXLH0XVxyyTvT/KOI6wLACwYCSU2RHd/vrvPTfLPkvx6kjdW1a1y6F3jJPlShl9MD7hTkusz/PJ+ZZJdBxZU1S2S3GH8cGPzr0jy2SR3Hj1K8fwMv+RuhCPFelSjHzpvzDAB8+QkrxvdUU9VfVuSN2U4AtDJoy/dF43Fvtb5y2j7pSSvTPKMJHcYbX/J2PanVtXq+TuN3tO4Lyb5qdU/DLv7Ft39v9c47vcnuXOS541+nH05w0cmfmyCO/LANqOeWNtWrCfWeI/fTPLqJPevqvFEHgCwoCSU2BBV9aSq2tnd/5TkmlHxPyXZP/r3O1at/tok/76qTq+qW2d4p/j1oyb0b0zyqKr6/lEHqC/I0b/0n5jk60muraq7JfnpjXpfR4l1vS5I8q8zHBVt9ag9N0/ybRmeo+ur6uwkPzTBfg/8ENufDDtuzfDO82r/LMm/q6rjq+pHk9w9wx8j434rwwTRvxjt67aj9ddyXm7sXPWM0XTPJLdIcvYE8QPbiHriiLZaPXGQUWLsyUm+nENbkwEAC0prAjbKWUleUlW3zLC5/BMO9OVQVf8pyfur6vjRer+TYXP69yY5IcMm8D+bJN39mar62Qz75blVkt/IsH+Hbx3h2M/MsPPUZyf5eJLXJ3nIBr2vw8Y6gfcm+Zsk13X3Rw4UdvffVtW/y7Bj2m9L8pYkF653p919aVX91yQfyPDH2O9l+EjBah/KsDXRVzO8W/647j7ky3x3/9Hoh9DrRne0/ybDpNEfrF6vqk5I8vgkT+nuL48t+x8ZJpvest73AGwr6onD2zL1xJhrRo2frk/yySSPHnu8DgBYYKVeZzMbfXm9JsPHFP5y3vEAsLmoJwAA5sMjb2w6VfWoqrrlqG+NFyf5dJLl+UYFwGahngAAmD8JJTajczLsEPRLGTbDf4Im8gCsop6ALaqqfqeqrqqqSw6zvKrq/6mqy6vqU1V131nHCMCQR94AAIBNoaoelOTaJL/X3eOdyKeqHplhP2WPzHCU2Zd19/fONkoAEi2UAACATaK735vkr4+wyjkZJpu6uz+Y5HZVdcpsogNgNQklAABgUZya5Iur5veNygCYsR3zDmA97njHO/ZgMJh3GACbzkc/+tGvdvfOeccxb+oJgLVt53qiqvYk2ZMkt7rVrb7nbne725wjAth8bko9sRAJpcFgkL179847DIBNp6pW5h3DZqCeAFjbFqwnrkhy2qr5XaOyQ3T3+UnOT5Ldu3e3egLgUDelnvDIGwAAsCguTPKU0Whv90/yN9195byDAtiOFqKFEgAAsPVV1WuTPDjJHatqX5JfSXJ8knT3byW5KMMR3i5P8o0kPzGfSAGQUAIAADaF7j73KMs7ydNnFA4ARyChBGw5//AP/5B9+/bluuuum3coG+aEE07Irl27cvzxx887lIXhOgAAgOmZekKpqo5LsjfJFd39I1V1epLXJblDko8meXJ3//204wC2j3379uXEE0/MYDBIVc07nJusu3P11Vdn3759Of300+cdzsJwHQAAwPTMolPun0ty2ar5X0/y0u7+riRfS/LUGcQAbCPXXXdd7nCHO2yJJEKSVFXucIc7bKmWNrPgOgAAgOmZakKpqnYl+eEkvz2aryQPSfLG0SoXJHnMNGMAtqetkkQ4YKu9n1nZaudtq70fAAAW17RbKP1Gkmcn+afR/B2SXNPd14/m9yU5dcoxAMzdT/7kT+bSSy+ddxjMmesAAICtYmoJpar6kSRXdfdHj3H7PVW1t6r27t+/f4OjA7aTwSCp2rhpMJg8ht/+7d/OPe5xj41+a0zAdQAAABtnmi2UHpDk0VW1nGEn3A9J8rIkt6uqA52B70pyxVobd/f53b27u3fv3LlzimECW93KStK9cdPKypGP93d/93f54R/+4dz73vfOPe95z7z+9a/Pgx/84OzduzdJ8qpXvSp3uctdcuaZZ+ZpT3tanvGMZ8zgLOA6AACAjTO1hFJ3P6+7d3X3IMkTkvxZdz8xycVJHjda7bwkb55WDADz8Pa3vz3f/u3fnk9+8pO55JJLctZZZ92w7Etf+lJe9KIX5YMf/GDe//7357Of/ewcI2WaXAcAAGxlsxjlbdxzkvxCVV2eYZ9Kr5pDDABTc6973SvvfOc785znPCd//ud/ntve9rY3LPvwhz+cH/iBH8hJJ52U448/Pj/6oz86x0iZJtcBAABb2Y6jr3LTdfe7k7x79PoLSc6cxXEB5uEud7lLPvaxj+Wiiy7KL/3SL+WhD33ovENiDlwHAABsZfNooQSwpX3pS1/KLW95yzzpSU/Ks571rHzsYx+7Ydn97ne/vOc978nXvva1XH/99XnTm940x0iZJtcBAABb2UxaKAFsJ5/+9KfzrGc9Kze72c1y/PHH5xWveEWe+cxnJklOPfXUPP/5z8+ZZ56Zk046KXe7290OehSKrcN1AADAViahBGx5S0vDYd43cn9H8ohHPCKPeMQjDip797vffcPrH/uxH8uePXty/fXX57GPfWwe85jHbFxwHJbrAAAANo6EErDlLS/PO4KDveAFL8if/umf5rrrrssP/dAPSSTMiOsAAAA2joQSwIy9+MUvnncIbAKuAwAAFplOuQEAAACYiIQSzNpgMOzI5cA0GMw7IgA42HhddbhJHQYA25ZH3mDWVlaS7hvnN7KXYADYCON11eGowwBg29JCCQAAAICJSCgBTMHy8nLuec97HlI+GAzy1a9+dQ4RMQ+uAwAAtioJJWDrW29fIOud9BmyYapquao+XVWfqKq9o7KTquqdVfX50b+335CDuQ4AAGDDSCgBW9+BvkA2alpZWddhr7/++jzxiU/M3e9+9zzucY/LN77xjRuWffOb38zZZ5+dV77ylUmSF73oRbnrXe+aBz7wgTn33HO325DyP9jdZ3T37tH8c5O8q7vvnORdo/mbznUAAAAbRkIJYEo+97nP5Wd+5mdy2WWX5Ta3uU1+8zd/M0ly7bXX5lGPelTOPffcPO1pT8tHPvKRvOlNb8onP/nJvO1tb8vevXvnHPncnZPkgtHrC5I8Zo6x3GSuAwAAtiIJJYApOe200/KABzwgSfKkJz0p73vf+5Ik55xzTn7iJ34iT3nKU5Ik73//+3POOefkhBNOyIknnphHPepRc4t5DjrJn1TVR6tqz6js5O6+cvT6y0lOnk9oG8N1AADAViShBDAlNTac9oH5BzzgAXn729+eXs+Q3FvfA7v7vknOTvL0qnrQ6oU9PElrnqiq2lNVe6tq7/79+2cQ6rFxHQAAsBVJKAFMyV/91V/lAx/4QJLkNa95TR74wAcmSV74whfm9re/fZ7+9KcnGSYW3vKWt+S6667Ltddem7e+9a1zi3nWuvuK0b9XJfmjJGcm+UpVnZIko3+vOsy253f37u7evXPnzlmFPDHXAQAAW5GEEsCU3PWud83LX/7y3P3ud8/Xvva1/PRP//QNy172spflm9/8Zp797Gfnfve7Xx796Efnu7/7u3P22WfnXve6V25729vOMfLZqKpbVdWJB14n+aEklyS5MMl5o9XOS/Lm+US4MVwHAABsRTvmHQDA1C0tDYd538j9HcVgMMhnP/vZQ8qXl5dveP27v/u7N7x+5jOfmRe84AX5xje+kQc96EH5nu/5ng0JdZM7OckfjR4B25HkNd399qr6SJI3VNVTk6wkefyGHM11AAAAG0ZCCdj6Vv1436z27NmTSy+9NNddd13OO++83Pe+9513SFPX3V9Icu81yq9O8tANP6DrAAAANoyEEsAm8JrXvGbeIbAJuA4AAFgU+lACAAAAYCISSsCWtNWGYt9q72dWttp522rvBwCAxSWhBGw5J5xwQq6++uot8+O7u3P11VfnhBNOmHcoC8V1AAAA06MPJWDL2bVrV/bt25f9+/fPO5QNc8IJJ2TXrl3zDmOhuA4AAGB6JJSALef444/P6aefPu8wmDPXAQAATI9H3gAAAACYiIQSAAAAABORUALmbjBIqg6eBoN5RwUAAMDh6EMJmLuVlWR8IK6q+cQCAADA0WmhBAAAAMBEJJQAAAAAmIiEEgAAAAATkVACAAAAYCISSgAAAABMREIJAAAAgIlIKAEAAAAwEQklAAAAACYiobTZDQZJ1Y3TYDDviAAAAIBtbse8A+AoVlaS7hvnq+YXCwCw2AaD4XeLo1lamnooAMBik1ACANguxm9UAQAcI4+8AQAAADARCSUAAAAAJiKhBAAAAMBEJJQAAAAAmMjUEkpVdUJVfbiqPllVn6mqXx2Vv7qq/rKqPjGazphWDAAAAABsvGmO8vatJA/p7mur6vgk76uqt42WPau73zjFYwMAAAAwJVNrodRD145mjx9NxqkFgC1sMEiqjjwNBvOOkk1hrYvFxQEAC2OqfShV1XFV9YkkVyV5Z3d/aLToP1XVp6rqpVX1bdOMAQCYnZWVpPvI08rKvKNkU1jrYnFxAMDCmGpCqbv/sbvPSLIryZlVdc8kz0tytyT3S3JSkuestW1V7amqvVW1d//+/dMMEza/8bu4C34Hd/ztLC3NOyIAAAAmMZNR3rr7miQXJzmru68cPQ73rSS/m+TMw2xzfnfv7u7dO3funEWYsHmN38Vd8Du4429neXneEQEAADCJaY7ytrOqbjd6fYskD0/y2ao6ZVRWSR6T5JJpxQAAAADAxpvmKG+nJLmgqo7LMHH1hu5+a1X9WVXtTFJJPpHk304xBgAAAAA22NQSSt39qST3WaP8IdM6JgAAAADTN5M+lAAAAADYOiSUAIBDbLHBJQEA2GDT7EMJAFhQB0ZjPKBqfrEAALD5aKEEAAAAwEQklAAAAACYiIQSAAAAABORUAIAAABgIhJKAAAAAExEQgkAAACAiUgoAQAztbSUVB08DQbzjmrBDQZOKgAwUzvmHQAAsL0sLx9aVjXzMLaWlZWk++AyJxUAmCItlAAAAACYiIQSAAAAABORUAIAAABgIhJKAAAAAExEQgkAANg0quqsqvpcVV1eVc9dY/mdquriqvp4VX2qqh45jzgBtjsJJbaF8dGUjaQMALD5VNVxSV6e5Owk90hyblXdY2y1X0ryhu6+T5InJPnN2UYJQJLsmHcAMAvjoykbSRkAYFM6M8nl3f2FJKmq1yU5J8mlq9bpJLcZvb5tki/NNEIAkmihBAAco/HWn1XJ0tK8o2KmlpYOvQg0A+amOTXJF1fN7xuVrfaCJE+qqn1JLkrys7MJDYDVJJQAgGNyoPXn6ml5ed5RMVPLy4deBCsr846Kre/cJK/u7l1JHpnkf1TVIb9rqmpPVe2tqr379++feZAAW52EEgAAsFlckeS0VfO7RmWrPTXJG5Kkuz+Q5IQkdxzfUXef3927u3v3zp07pxQuwPYloQQAAGwWH0ly56o6vapunmGn2xeOrfNXSR6aJFV19wwTSpogAcyYhBIAALApdPf1SZ6R5B1JLstwNLfPVNULq+rRo9V+McnTquqTSV6b5Me7Vw+/AsAsGOUNAADYNLr7ogw7215d9surXl+a5AGzjguAg2mhBAAAAMBEJJQAAAAAmIiEEgAAm8PSUlJ18DQYzDsqAGANEkpsGoPB7L4/+r4KAJvQ8nLSffC0sjLvqACANeiUm01jZWX4vfGAqukda3n50LJpHg8AAAC2Ei2UAAAAAJiIhBIAAAAAE5FQAgAAAGAiEkoAAAAATERCCQAAAICJSCgBwDY3GAxHulw9LS3NOyoAADazHfMOAACYr5WVpHveUQAAsEi0UAIAAABgIhJKAAAAAExEQmnRrdXxxWAw76imZvztbuG3yhp8/ltTVR1XVR+vqreO5k+vqg9V1eVV9fqquvm8YwQAAA4mobToDnR8sXpaWZl3VFMz/na38FtlDT7/Levnkly2av7Xk7y0u78rydeSPHUuUQEAAIcloQTA3FTVriQ/nOS3R/OV5CFJ3jha5YIkj5lPdAAAwOFIKAEwT7+R5NlJ/mk0f4ck13T39aP5fUlOnUdgAADA4UkoATAXVfUjSa7q7o8e4/Z7qmpvVe3dv3//BkcHAAAciYQSAPPygCSPrqrlJK/L8FG3lyW5XVXtGK2zK8kVa23c3ed39+7u3r1z585ZxAsAAIxMLaFUVSdU1Yer6pNV9Zmq+tVRudF7AEh3P6+7d3X3IMkTkvxZdz8xycVJHjda7bwkb55TiAAAwGFMs4XSt5I8pLvvneSMJGdV1f1j9B4Ajuw5SX6hqi7PsE+lV805HgAAYMzUEko9dO1o9vjR1DF6DwBjuvvd3f0jo9df6O4zu/u7uvtHu/tb844PAAA42FT7UKqq46rqE0muSvLOJH8Ro/cAAAAALLSpJpS6+x+7+4wMO1U9M8nd1rut0XtYj6WlpOrgaTCYYQCDwcYffHyfVcM3yrpM4yMBAADgYDuOvspN193XVNXFSb4vo9F7Rq2Ujjh6T5Lzk2T37t09izhZPMvLh5ZVzTCAlZWkV12eG3Hw8X0ykWl8JAAAABxsmqO87ayq241e3yLJw5NcFqP3AMDMrNXo8VgaQa7VIlTjSQCA7WuaLZROSXJBVR2XYeLqDd391qq6NMnrquo/Jvl4jN4DAFOzUY0e12oRCgDA9jW1hFJ3fyrJfdYo/0KG/SkBAAAAsICm2ik3AAAAAFuPhBIAAAAAE5FQAgAAAGAiEkoAAAAATERCCQAAAICJSCgBAHO3tJRU3TgNBvOOiA01GBz8AVcNP3QAYGHtmHcATMGBb+Wr55eX5xYOR7HW58W6jJ+6A2XA4hmvpsb/b7PgVlaS7nlHAQBsIAmlrci38sUi2XfMnDoAAID58MgbAAAAABORUAIAAABgIhJKAAAAAExEQgkAAACAiUgoAQAAADARCSUAYGENBsPBTA9Mg8G8IwIA2B52zDsAAIBjtbKSdN84XzW/WAAAthMtlAAAAACYiIQSQ54Z2BhLSwedx+UaOJUAAABsOR55Y8gzAxtjeTlVN57KQVVWVuYbEgAAAGw0LZQAAAAAmIiEEgAAAAATkVACAAAAYCISSgDApjM2xsHWGy9ifDCMLfcGN9CWvxgAYDHplBsA2HSWlw8t21LjRYwPhpFssTe4gbb8xQAAi0kLJQAAAAAmIqEEAAAAwEQklAAAAACYiIQSAAAAABORUJqltUZ0Odq0tHTTjzs+OsqCjIyy1qAuG3E61nu8WZ+m8ctjQT4mAAAAtiGjvM3SWiO6zML46CgLMjLKWoO6zPJ4sz5N45fHgnxMAAAAbENaKAEAAAAwEQklAAAAACYioQQAAADARCSUAAAAAJiIhBIAAAAAE5FQAgAAAGAiEkoAAAAATERCCQAAAICJSCgBAAAAMBEJJSBJMhgkVTdOg8GxrbNRlpYOPlbVsAwYGv//OO3/k2wia3342+0P5FqVhP8AADBTO+YdALA5rKwk3TfOVx3bOhtleXl6+4atYPz/YzLd/5NsImt9+NvNWpWE/wAAMFNaKAEAAAAwEQklAAAAACYioQQAAADARCSUAAAAAJiIhBIAAAAAE5laQqmqTquqi6vq0qr6TFX93Kj8BVV1RVV9YjQ9cloxAAAAALDxdkxx39cn+cXu/lhVnZjko1X1ztGyl3b3i6d4bAAAAACmZGoJpe6+MsmVo9d/W1WXJTl1WscDAAAAYDZm0odSVQ2S3CfJh0ZFz6iqT1XV71TV7Q+zzZ6q2ltVe/fv3z+LMIFVlpaSqoOnpaV5RwUAAMBmMPWEUlXdOsmbkvx8d389ySuSfGeSMzJswfRf19quu8/v7t3dvXvnzp3TDhMYs7ycdB88LS/POyoAYCurqrOq6nNVdXlVPfcw6zx+VT+tr5l1jAAMTbMPpVTV8Rkmk36/u/8wSbr7K6uWvzLJW6cZAwAAsPlV1XFJXp7k4Un2JflIVV3Y3ZeuWufOSZ6X5AHd/bWq+mfziRaAaY7yVkleleSy7n7JqvJTVq322CSXTCsGAABgYZyZ5PLu/kJ3/32S1yU5Z2ydpyV5eXd/LUm6+6oZxwjAyDRbKD0gyZOTfLqqPjEqe36Sc6vqjCSdZDnJT00xBgAAYDGcmuSLq+b3JfnesXXukiRV9f4kxyV5QXe/fTbhAbDaNEd5e1+SWmPRRdM6JgAAsKXtSHLnJA9OsivJe6vqXt19zfiKVbUnyZ4kudOd7jTLGAG2hZmM8gYAsOUMBocOhzkYzDsqWGRXJDlt1fyuUdlq+5Jc2N3/0N1/meT/yzDBdAiD/ABMl4QSAMCxWFk5dDjMlZV5RwWL7CNJ7lxVp1fVzZM8IcmFY+v8rwxbJ6Wq7pjhI3BfmGWQAAxJKAEAAHPX3dcneUaSdyS5LMkbuvszVfXCqnr0aLV3JLm6qi5NcnGSZ3X31fOJGGB7m2an3AAAAOvW3RdlrM/V7v7lVa87yS+MJgDmSAslAAAAACYioTQta3XUubQ076g2jUU4PX+ZwYZ0tLq0dPBukhn12Tp+knUUewOnBgAA4KbxyNu0HOiokzUtwukZZCzIA9mgCS0vjxXUjPpsHT/Jxxj/VuTUAAAA3DRaKAEAAAAwEQklAGAhjD9CPNPHpRfhWW0AgBnyyBsAsBAOeYR4lhbhWW0AgBnSQgkAAACAiUgoAQAAADARCSUA5qKqTqiqD1fVJ6vqM1X1q6Py06vqQ1V1eVW9vqpuPu9YAQCAg0koATAv30rykO6+d5IzkpxVVfdP8utJXtrd35Xka0meOscYAQCANUgoATAXPXTtaPb40dRJHpLkjaPyC5I8Zg7hAQAARyChBMDcVNVxVfWJJFcleWeSv0hyTXdfP1plX5JT5xUfAACwNgkltpylpWS5BknVDdMXjxsc075W7SJJMji23WyYwWAYy3KWDg5uwsAO7Gf1tLQ0jYjnb2lp+7zXRdTd/9jdZyTZleTMJHdb77ZVtaeq9lbV3v37908tRhbfWn/zxv9srmedDQ1gvX+IFvEPtj+8ALAt7Jh3ALDRlpeT1ErSfUPZrgMZoQmt2kVSycrKTQrtJlu54W0tJxl+R+/OjRmvifez9S0vzzsC1qO7r6mqi5N8X5LbVdWOUSulXUmuOMw25yc5P0l27969Ta5ojsVaf/PG/2yuZ50NDWAW284V/TZJAAAgAElEQVSLP7wAsC1ooQTAXFTVzqq63ej1LZI8PMllSS5O8rjRauclefN8IgQAAA5HCyUA5uWUJBdU1XEZ3uB4Q3e/taouTfK6qvqPST6e5FXzDBIAADiUhBIAc9Hdn0pynzXKv5Bhf0oAAMAm5ZE3AAAAACYioQQAAADARCSUAAAAAJiIhBIAwDoNBknVjdNgsIE7X1o6eOdVw7LNsj8AgFV0yg0AsE4rK0n3jfNVG7jz5eUN3NkU9gcAsIoWShtl/Jblot8BHL+ruaG3YFeZ6q3eTWCL3x1e6+1ttY8QAACAQ03cQqmqbp/ktNFwzxwwfsty0Y3f1dzQW7CrTPVW7yZwuLvDW+RtrvX2ttpHyOTUEwAAsPWtq4VSVb27qm5TVScl+ViSV1bVS6YbGgCLQj0BAADby3ofebttd389yb9M8nvd/b1JHja9sABYMOoJAADYRtabUNpRVackeXySt04xHgAWk3oCAAC2kfUmlH41yTuSXN7dH6mq70jy+emFBcCCUU8AAMA2st5Oua/s7u8+MNPdX9A3BgCrqCcAAGAbWW8Lpf93nWUAbE/qCQAA2EaO2EKpqr4vyfcn2VlVv7Bq0W2SHDfNwADY/NQTAACwPR3tkbebJ7n1aL0TV5V/PcnjphUUAAtDPQEAANvQERNK3f2eJO+pqld398qMYgJgQagnAABge1pvp9zfVlXnJxms3qa7HzKNoDiCpaWk6uD5rWT8/R0oW16eSzgHWVpKr1RSB5dtCoc7b5sglM3y8TF16gkAANhG1ptQ+oMkv5Xkt5P84/TC4ai2+i/ztd7feKJkXpaXU5V0zzuQNWyi62I8lM3y8TF16gkAANhG1ptQur67XzHVSABYZOoJAADYRm62zvXeUlU/U1WnVNVJB6apRgbAIlFPAADANrLeFkrnjf591qqyTvIdGxsOAAtKPbEJbKLu1AAA2OLWlVDq7tOnHQgAi0s9sTlsou7UAADY4taVUKqqp6xV3t2/t7HhALCI1BMAALC9rPeRt/uten1Ckocm+ViSw/5QqKrTRstPzvCxh/O7+2WjPjVen+HQ0stJHt/dX5s4cgA2k4nrCQAAYHGt95G3n109X1W3S/K6o2x2fZJf7O6PVdWJST5aVe9M8uNJ3tXdv1ZVz03y3CTPmThyADaNY6wnAACABbXeUd7G/V2SI/aX0d1XdvfHRq//NsllSU5Nck6SC0arXZDkMccYAwCb11HrCQAAYHGttw+lt2T42FqSHJfk7knesN6DVNUgyX2SfCjJyd195WjRlzN8JG6tbfYk2ZMkd7rTndZ7KNbDMEBrGwySlZUb59dxTsY3WedmsOXc1HoCAABYLOvtQ+nFq15fn2Slu/etZ8OqunWSNyX5+e7+eq1KZHR3V1WvtV13n5/k/CTZvXv3mutwjAwDtLaVlaQnu9SOYRPYqo65ngAAABbPuh556+73JPlskhOT3D7J369nu6o6PsNk0u939x+Oir9SVaeMlp+S5KpJgwZgcznWegK2nAOtgFdPmq4CAFvQuhJKVfX4JB9O8qNJHp/kQ1X1uKNsU0leleSy7n7JqkUXJjlv9Pq8JG+eNGgANpdjqSdgS1peHjZdXT1pGQwAbEHrfeTtPyS5X3dflSRVtTPJnyZ54xG2eUCSJyf5dFV9YlT2/CS/luQNVfXUJCsZ/vAAYLEdSz0BAAAsqPUmlG524EfCyNU5Suum7n5fkjrM4oeu87gALIaJ6wkAAGBxrTeh9PaqekeS147m/3WSi6YTEgALSD0BAADbyBETSlX1XUlO7u5nVdW/TPLA0aIPJPn9aQcHwOamngAAgO3paC2UfiPJ85JkNErbHyZJVd1rtOxRU40OgM1OPQEAANvQ0fq3OLm7Pz1eOCobTCUiABaJegIAALaho7VQut0Rlt1iIwMBYCGpJ1hIS0tJ1cHzAACs39FaKO2tqqeNF1bVTyb56HRCYmENBsNv50eaBoPZHGcL/DI48GPnpr6tA6crmc/p2aj3waalnmAhLS8n3TdOy8vzjggAYLEcrYXSzyf5o6p6Ym78YbA7yc2TPHaagbGAVlaG38qPZPXt4GkeZwvYqB83N5yums9p8yNty1NPAADANnTEhFJ3fyXJ91fVDya556j4j7v7z6YeGQCbnnoCAAC2p6O1UEqSdPfFSS6eciwALCj1BAAAbC9H60MJAAAAAA4ioQQAAADARCSUAAAAAJiIhBIAAAAAE5FQAgAAAGAiEkoAAAAATERCCQAAAICJ7Jh3ADATS0tJ1dHXAQAAAI5KQontYXl53hEAAADAluGRNwAAAAAmIqEEAAAAwEQklABgAQwGw67gDky6fdsg4ye2ali2vUMBADgqfSgBwAJYWUm65x3FFrTWiT3aIA5bPxQAgKPSQgkAAACAiUgoAQAAADARCSUAAAAAJiKhBAAAAMBEJJSOleF2JnLgdCXHcLqWljbNud5EobCBxj9XIysBAAAcmVHejpXhdiZyw+mqYzhty8tTiOjYbKJQ2EBrfa5GVgIAADg8LZQAgC3jQIvD5Rrc0OSwU4c2Q1xHU9PVrRcTLVMBAFaTUAIAtozl5WFL2EFGTWOPNh2h6emBfR1oWXuU1YENUlVnVdXnquryqnruEdb7V1XVVbV7lvEBMCShBAAAbApVdVySlyc5O8k9kpxbVfdYY70Tk/xckg/NNkIADpBQAgAANoszk1ze3V/o7r9P8rok56yx3ouS/HqS62YZHAA3klACAAA2i1OTfHHV/L5R2Q2q6r5JTuvuP55lYAAcTEIJAABYCFV1syQvSfKL61h3T1Xtraq9+/fvn35wANuMhBIAALBZXJHktFXzu0ZlB5yY5J5J3l1Vy0nun+TCtTrm7u7zu3t3d+/euXPnFEMG2J4klAAAgM3iI0nuXFWnV9XNkzwhyYUHFnb333T3Hbt70N2DJB9M8uju3jufcAG2LwklAIANNBgkVQdPg8G8o9oGlpac+C2gu69P8owk70hyWZI3dPdnquqFVfXo+UYHwGo75h0A28yBL3ur5wFgC1lZSboPLltd9TEly8uHljnxC6m7L0py0VjZLx9m3QfPIiYADiWhxGyt9WUPAAAAWCgeeQMAAABgIhJKAAAAAExEQgmAuaiq06rq4qq6tKo+U1U/Nyo/qareWVWfH/17+3nHCgAAHExCCYB5uT7JL3b3PZLcP8nTq+oeSZ6b5F3dfeck7xrNAwAAm4iEEgBz0d1XdvfHRq//NsPhoU9Nck6SC0arXZDkMfOJEAAAOJypJZSq6neq6qqqumRV2Quq6oqq+sRoeuS0jg/A4qiqQZL7JPlQkpO7+8rRoi8nOfkw2+ypqr1VtXf//v0ziXMaBoPhyOarp8Fg3lEBAMCRTbOF0quTnLVG+Uu7+4zRdNEUjw/AAqiqWyd5U5Kf7+6vr17W3Z2k19quu8/v7t3dvXvnzp0ziHQ6VlaS7oOnlZV5RwUAAEc2tYRSd783yV9Pa/8ALL6qOj7DZNLvd/cfjoq/UlWnjJafkuSqecUHAACsbR59KD2jqj41eiTOyD0A21RVVZJXJbmsu1+yatGFSc4bvT4vyZtnHRsAAHBks04ovSLJdyY5I8mVSf7r4VbcKn1jbGnjHX8sLc07ooWxtHRonykbdfp8LCyQByR5cpKHjPWt92tJHl5Vn0/ysNE8AACwieyY5cG6+ysHXlfVK5O89Qjrnp/k/CTZvXv3mv1nMGcHOv5gYsvL09u3j4VF0d3vS1KHWfzQWcYCAABMZqYtlA70iTHy2CSXHG5dAAAAADanqbVQqqrXJnlwkjtW1b4kv5LkwVV1RoYj9iwn+alpHR8AAACA6ZhaQqm7z12j+FXTOh4AAAAAszGPUd4AALaEYx1kYXwABYMoAACLZqadcgMAbCXHOsiCARQAgEWnhRIAAAAAE5FQAgAAAGAiEkoAAAAATERCCQAAAICJSCgBAAAAMBEJJQAAAAAmsmPeAQAAbCpLS0nVoWUAANxAQgkAYLXl5XlHAACw6XnkDQAAAICJSCgBAAAAMBEJJQAAAAAmIqEEAAAAwEQklAAAAACYiIQSAAAAABORUGJ9lpaSqoOnpaV5R7Xt+VgA4AjWqigHg3lHBQBbwo55B8CCWF6edwSswccCAEewVkVZNfMwAGAr0kIJAAAAgIlIKAEAAAAwEQklAAAAACYioQQAAADARCSUAAAAAJiIhBIAAAAAE5FQAgAAAGAiEkoAAAAATERCCQAAAICJSCgBAAAAMBEJJQAAAAAmIqEEAAAAwEQklAAAAACYiIQSAAAAABPZMe8AAICDLS0lVYeWAQDAZiGhBACbzPLyvCMAAIAj88gbAAAAABORUAIAAABgIhJKAAAAAExEQokNNxgMO5NdPelMlkVzoFPkI02DwbyjBAAAmA+dcrPhVlaS7nlHATfNejpFHh+FCwAAYLvQQgkAAACAiUgoAcCUrPUIsEclt6fxx2jX+yj4+HZrXT+uMwBgHjzyBgBTstYjwB6V3J7W8xjterZb6/pxnQEA86CFEgAAAAATkVACAAAAYCJTSyhV1e9U1VVVdcmqspOq6p1V9fnRv7ef1vEBAAAAmI5ptlB6dZKzxsqem+Rd3X3nJO8azQMAAACwQKaWUOru9yb567Hic5JcMHp9QZLHTOv4AAAAAEzHrPtQOrm7rxy9/nKSk2d8fAAAAABuorl1yt3dnaQPt7yq9lTV3qrau3///hlGtobBYDj+7uppaWm+MW0i46fHqdmklpYO/qAGg3lHBLAx1NMAADO3Y8bH+0pVndLdV1bVKUmuOtyK3X1+kvOTZPfu3YdNPM3EykrS8w1hM3N6FsTy8sHzVXMJA2DDqYgAAGZu1i2ULkxy3uj1eUnePOPjAwAAAHATTS2hVFWvTfKBJHetqn1V9dQkv5bk4VX1+SQPG80DAAAAsECm9shbd597mEUPndYxAQAAAJi+uXXKDQAAAMBiklACAAAAYCISSgAAAABMREIJAIDtY2kpqTp4GgzmHRUALJypdcoNAACbzvLyoWVVMw8DABadFkprGQwOvmu1tDTviAAAAAA2DS2U1rKyknTPOwoAAACATUkLJQAAAAAmIqEEAAAAwEQklAAAAACYiIQSAACwKVTVWVX1uaq6vKqeu8byX6iqS6vqU1X1rqoyeg7AnEgoAQAAc1dVxyV5eZKzk9wjyblVdY+x1T6eZHd3f3eSNyb5L7ONEoADJJQAAIDN4Mwkl3f3F7r775O8Lsk5q1fo7ou7+xuj2Q8m2TXjGAEYkVACAAA2g1OTfHHV/L5R2eE8NcnbphoRAIe1Y94BAAAATKKqnpRkd5IfOMI6e5LsSZI73elOM4oMYPvQQgkAANgMrkhy2qr5XaOyg1TVw5L8hySP7u5vHW5n3X1+d+/u7t07d+7c8GABtjsJJQDmoqp+p6quqqpLVpWdVFXvrKrPj/69/TxjBGCmPpLkzlV1elXdPMkTkly4eoWquk+S/55hMumqOcQIwIiEEgDz8uokZ42VPTfJu7r7zkneNZoHYBvo7uuTPCPJO5JcluQN3f2ZqnphVT16tNr/neTWSf6gqj5RVRceZncATJk+lACYi+5+b1UNxorPSfLg0esLkrw7yXNmFhQAc9XdFyW5aKzsl1e9ftjMgwJgTVooAbCZnNzdV45efznJyfMMBgAAWJuEEgCbUnd3kj7c8qraU1V7q2rv/v37ZxjZTbO0lFTdOC0tzTsi4JD/mFXJYDDvqABgU/PIGwCbyVeq6pTuvrKqTkly2A5Xu/v8JOcnye7duw+beNpslpfnHQFwiLX+Y1bNPAwAWCRaKAGwmVyY5LzR6/OSvHmOsQAAAIchoQTAXFTVa5N8IMldq2pfVT01ya8leXhVfT7Jw0bzAADAJuORNwDmorvPPcyih840EAAAYGJaKAEAAAAwEQklAAAAACYioQQAAADARPShBACwIJaWDh3NfmlpPrEAANubhBIAwIJYXp53BAAAQx55AwAAAGAiEkoAAAAATERCCQAAAICJSCgBAAAAMBEJJQAAAAAmIqEEAAAAwEQklACA+RsMkqqDp8FgfestLc042M1vaenop5KjGD+JTiQAHGTHvAPYFAaDZGXlxnlfTA9r/FQlThfb14HfGuNly8tzCQcW28pK0n1w2fh/sMOtxyHG/w6tdSo5irX+mDuRAHADCaXEl9MJOFVwI781AACA7cojbwAAAABMREIJAAAAgIlIKAEAAAAwEQklAAAAACYioQQAAADAROYyyltVLSf52yT/mOT67t49jzgAAAAAmNxcEkojP9jdX53j8QEAAAA4Bh55AwAAAGAi80oodZI/qaqPVtWeOcUAAAAAwDGYV0Lpgd193yRnJ3l6VT1ofIWq2lNVe6tq7/79+2cfIWx1S0tJ1cHTYHDkbQaDybcBAABgy5lLQqm7rxj9e1WSP0py5hrrnN/du7t7986dO2cdImx9y8tJ98HTysqRt1lZmXwbAAAAtpyZJ5Sq6lZVdeKB10l+KMkls44DAAAAgGMzj1HeTk7y/7d377G2lOUdx78/PSIRES94QaBnS4oX0FiREPGKl1ikidSIFlqjprRUG2tt/yJpY41NLzRRkrZqa9V4qQItrUpbvIEgrSkoKlcRRHqOglS8IC0aUevTP9ZsWKyz99mz9l5rZtZe30+ycmbPmpn1zHNm3nfWu+Z958NJVj//Q1X18R7ikCRJkiRJ0iZ03qBUVTcBT+76cyVJ0oJZHettcp7mZmXl3j2Zd+4c9ZBWY71j0iRJkpZQH3coSZIkbcwv6Z1bHSpv1WTbydJb65g0SZKkJdXXU94kSZIkSZK0oGxQkiRJkiRJ0lRsUJIkSZIkSdJUbFCSJEmSJEnSVGxQ0rpWVkbjTI6/fLjOkpk8CDwAJEmSJEn4lDftxeSTXrSEPAgkSZIkSWvwDiVJkjQ/a93uutbLOyAlSZIWincoSZKk+fFOR0mSpG3JO5QkSZIkSZI0FRuUJEmSJEmSNBUblCRJkiRJkjQVG5QkSZIkSZI0FRuUJEmSJEmSNBUblCRJkiRJkjQVG5QkSdqElRVI9v7aubPvKKWRnTv3PD5XVvqOSpIkLbLt36DkFf+a1kqLF5aS1N7u3VC199euXX1HKY3s2rXn8bl7d99RSZKkRbaj7wDmbvWKX/eyVlqSfmKRJEmSJEmLZfvfoSRJkiRJkqSZskFJkiRJkiRJU7FBSZIkSZIkSVOxQUmSJEmSJElTsUFJkiRJkiRJU7FBSZKkMSsro6dejr9WVvqOSpq9nTv3PNZ37uw7KkmStCh29B2AJElDsns3VN17XtJPLNI87drVdwSSJGmReYeSJEmSJEmSpmKDkiRJkiRJkqZig5Kke0wOqOFgGlObTKFj70iSJEnajhxDSdI9HFBjyyZT6Ng7kiRJkrYj71CSJEmSJEnSVGxQkiRJkiRJ0lRsUJIkSZIkSdJUbFCSJC21lRXHopfWM/mggfUeNjB5HvlAAkmStj8H5ZYkLbXdu6Gq7yikYVrrWQ1rPWxg8jzygQSSJG1/3qEkSZIkSZKkqdigJEmSJEmSpKnYoCRJkiRJkqSp2KCku00OvOnAtNLWrTWgbZsBbiVJkiRpyByUW3dba+BNSVvT5rxy8FpJkiRJi8Y7lCRJkiRJkjQVG5QkSdrAWl0XB9EteGWlXR/KrSy31dcgEqVZanM+rLVMm+69bQ/VQWnTt3maPs8LmQRJ0jKyy5skSRsYbJfg3buh6t7z1upDuZXlpAltzoe1lmnTvbftoTooWykgtnK+SpLUM+9QkiRJkiRJ0lRsUJIkSZIkSdJUemlQSnJ8kuuT3Jjk9D5ikCQNl/WEJC2vjeqAJPdPck7z/mVJVrqPUpLUeYNSkvsCbwNeBBwBnJLkiK7jkCQNk/WEJC2vlnXAqcDtVfXzwJnAGd1GKUmCfu5QOga4sapuqqofA2cDJ/YQhyRpmKwnJGl5takDTgTe10yfCzw/ceRySepaHw1KBwPfGPv75maeJElgPSFJy6xNHXD3MlX1U+AO4GGdRCdJutuOvgNYT5LTgNOaP+9Mcn3LVQ8EvjOxsRlGtml7xtWzZHgxNYyrvWHEtOc5Noy47m2IMUET1xaKqZ2zC2WxbKGeGLfV/PdvreDXOieTPY//dusuiqGe47O20PvZ5vBa6/pkcQ/LvRrtZ9vzcPNJsJ4YuSvJNX3GMwALXX7MkHkYMQ/mYNXjNrtiHw1KtwCHjv19SDPvXqrqncA7p914ksur6ujNhzcfQ4xriDGBcU1jiDHBMOMaYkww3Lh6Ntd6Ytyy5H8Z9nMZ9hGWYz+XYR9hefZzE9rUAavL3JxkB3AA8N3JDY3XE+bbHKwyDyPmwRysSnL5Ztfto8vb54HDkzwmyT7AycB5PcQhSRom6wlJWl5t6oDzgFc10ycBn66q6jBGSRI93KFUVT9N8jrgE8B9gfdU1bVdxyFJGibrCUlaXuvVAUneDFxeVecB7wY+kORG4HuMGp0kSR3rZQylqjofOH9Om99S94c5GmJcQ4wJjGsaQ4wJhhnXEGOC4cbVqznXE+OWJf/LsJ/LsI+wHPu5DPsIy7OfU1urDqiqN45N/wh42ZSbNd/mYJV5GDEP5mDVpvMQ7w6VJEmSJEnSNPoYQ0mSJEmSJEkLbOEblJK8LMm1SX6WZN0R2pPsSnJ1kiu2Mor5HOI6Psn1SW5McvqcY3pokk8l+Wrz70PWWe7/mjxdkWRuA+FutO9J7p/knOb9y5KszCuWKWJ6dZJvj+XnNzqI6T1JblvvUbcZ+csm5quSHDXvmFrGdVySO8Zy9ca1lptxTIcmuSjJl5vz73fXWKbTfLWMqfNcLashls3zMLTyfpaGWHfM2hDrolkbat02a0OsK7ezZSgf2miRh99vrkuuSnJhkp19xDlvbevyJC9NUnu7LlhUbXKQ5OVj16kf6jrGLrQ4J36uuV7/UnNenNBHnPM0t3q3qhb6BTwBeBxwMXD0XpbbBRw4pLgYDTT4NeAwYB/gSuCIOcb0F8DpzfTpwBnrLHdnB/nZcN+B3wb+ppk+GThnADG9Gvjrro6j5jOfDRwFXLPO+ycAHwMCPA24bCBxHQf8a8e5Ogg4qpneH7hhjf/DTvPVMqbOc7WsryGWzXPaz8GU9zPer8HVHT3tY+d10Rz2c5B1Ww/7afk/u1xv+/Jhhnl4LvCAZvq1y5qHZrn9gUuAS9e7LljUV8tj4XDgS8BDmr8f0XfcPeXhncBrm+kjgF19xz2HPMyl3l34O5Sq6rqqur7vOCa1jOsY4MaquqmqfgycDZw4x7BOBN7XTL8P+OU5ftZG2uz7eLznAs9Pkp5j6lxVXcLoCSbrORF4f41cCjw4yUEDiKtzVXVrVX2xmf5f4Drg4InFOs1Xy5jUkYGWzfMwpPJ+loZYd8zadjj+NjTUum3WhlhXbmPLUD60sWEequqiqvph8+elwCEdx9iFtmXpHwNnAD/qMriOtMnBbwJvq6rbAarqto5j7EKbPBTwoGb6AOCbHcbXiXnVuwvfoDSFAj6Z5AtJTus7mMbBwDfG/r6Z+X7RfGRV3dpM/zfwyHWW2zfJ5UkuTTKvLyFt9v3uZarqp8AdwMPmFE/bmABe2twGeG6SQ+cYT1tdH0fTODbJlUk+luTILj+4uY39KcBlE2/1lq+9xAQ95kp7GPI51daQyvtZGmLdMWuLWhfN2nY4D9uy/J+NZSgf2pj23DmV0V0J282GeWi69BxaVf/WZWAdanMsPBZ4bJLPNtcCx3cWXXfa5OFNwCuS3MzoCZO/001og7KpenfH3MKZoSQXAI9a460/qKqPttzMM6vqliSPAD6V5CtNK13fcc3U3mIa/6OqKsl6j/jb2eTqMODTSa6uqq/NOtYF9S/AWVV1V5LfYvQr1/N6jmmovsjoWLqz6Yf8EUa31c5dkgcC/wS8oar+p4vP3MgGMfWWq+1oiGXzPFjeLzXrou3D8l+9SfIK4GjgOX3H0rUk9wHeyqgL8TLbwajMOY7RnWqXJHlSVX2/16i6dwrw3qp6S5JjgQ8keWJV/azvwIZuIRqUquoFM9jGLc2/tyX5MKNb37bUoDSDuG4Bxn9VPKSZt2l7iynJt5IcVFW3NrevrXlL41iubkpyMaM7Kmb9BaPNvq8uc3OSHYxuP/zujOOYKqaqGv/8dzEap6RvMz+OZmG80aSqzk/y9iQHVtV35vm5Se7HqOHmg1X1z2ss0nm+Noqpr1xtV0Msm+dhgcr7WRpi3TFri1oXzdpCnIdbZfk/U8tQPrTR6txJ8gJGP0A8p6ru6ii2Lm2Uh/2BJwIXN70eHwWcl+TFVTX3Bzh1pM2xcDOjsXJ+AvxXkhsYNTB9vpsQO9EmD6cCxwNU1X8m2Rc4kHWun7apTdW7S9HlLcl+SfZfnQZeCKw5unnHPg8cnuQxSfZhNDjgPJ+ycx7wqmb6VcAev9QneUiS+zfTBwLPAL48h1ja7Pt4vCcBn66q9X5l7ySmiX6kL2Y0Hk7fzgNe2YzM/zTgjrGuLr1J8qjVcQmSHMOovJnrRVvzee8Grquqt66zWKf5ahNTH7nSXnVdNs/DkMr7WRpi3TFri1oXzdog67ZZs/yfqWUoH9poU4Y8Bfhb4MXbdMwc2CAPVXVHVR1YVStVtcJoLKnt1JgE7c6JjzC6O2n1WuCxwE1dBtmBNnn4OvB8gCRPAPYFvt1plP3bXL1bAxhxfCsv4CWMWlbvAr4FfKKZ/2jg/Gb6MEajuV8JXMuo20PvcdU9o6nfwOgX4bnGxaiP+IXAV4ELgIc2848G3tVMPx24usnV1cCpc4xnj30H3syoMIfRifyPwI3A54DDOvh/2yimP2uOoSuBi4DHdxDTWcCtwE+aY+pU4DXAa5r3A7ytiflqOnpCRYu4XjeWq0uBp3cQ0zMZjZd2FXBF8zqhz3y1jKnzXBaaGRcAAAQKSURBVC3ra4hl85z2c1Dl/Yz3bXB1Rw/72HldNId9HGTd1sN+Wv7PNt/bvnyYUR4uaOrA1euS8/qOuY88TCx78aKWM1s8FsKo69+Xm7L25L5j7ikPRwCfbcriK4AX9h3zHHIwl3o3zcqSJEmSJElSK0vR5U2SJEmSJEmzY4OSJEmSJEmSpmKDkiRJkiRJkqZig5IkSZIkSZKmYoOSJEmSJEmSpmKDkraNJBcl+cWJeW9I8o4pt3N+kgdvsMyd68x/b5KTpvm8Zr0rkpw97XqSpPYWsZ5I8qYktzT1xFeSvCOJ12+SJKl3XpBoOzkLOHli3snN/A1l5D5VdUJVfX/m0a3/uU8A7gs8K8l+XX2uJC2hhawngDOr6heAI4AnAc/p8LMlSZLWZIOStpNzgV9Ksg9AkhXg0cC/J3lgkguTfDHJ1UlOXF0myfVJ3g9cAxyaZFeSA5v3P5LkC0muTXLa+IclObOZf2GSh08Gk+SpST7TrP+JJAetE/cpwAeATwInziQTkqS1LGo9sWofYF/g9q2lQZIkaetsUNK2UVXfAz4HvKiZdTLwD1VVwI+Al1TVUcBzgbckSbPc4cDbq+rIqto9sdlfr6qnAkcDr0/ysGb+fsDlVXUk8Bngj8ZXSnI/4K+Ak5r13wP8yTqh/wpwNqNfyE/ZxK5LklpY4Hri95JcAdwK3FBVV2xm/yVJkmbJBiVtN+PdGca7MQT40yRXARcABwOPbN7bXVWXrrO91ye5ErgUOJTRlwqAnwHnNNN/DzxzYr3HAU8EPtV8CfhD4JDJjSc5GvhOVX0duBB4SpKHttxXSdL0FqqeaKx2eXsEsF+SyW57kiRJndvRdwDSjH0UODPJUcADquoLzfxfAx4OPLWqfpJkF6NuAwA/WGtDSY4DXgAcW1U/THLx2DqTanJ14NqqOnaDeE8BHt/EA/Ag4KXA322wniRpcxatnrhnA6O4Pg48m9GdrZIkSb3xDiVtK1V1J3ARo64D44OsHgDc1lyMPxfY2WJzBwC3N18SHg88bey9+wCrT+n5VeA/Jta9Hnh4kmNh1LUhyZHjCzRP6Xk58KSqWqmqFUZjKNntTZLmZJHqiUlNF7xnAF9rEZskSdJc2aCk7egs4Mnc+4vCB4Gjk1wNvBL4SovtfBzYkeQ64M8ZdWdY9QPgmCTXAM8D3jy+YlX9mNEXiTOarhBXAE+f2P6zgFuq6ptj8y4BjmgxMKskafMWpZ5YtTqG0jWMngr69haxSZIkzVVG41BKkiRJkiRJ7XiHkiRJkiRJkqZig5IkSZIkSZKmYoOSJEmSJEmSpmKDkiRJkiRJkqZig5IkSZIkSZKmYoOSJEmSJEmSpmKDkiRJkiRJkqZig5IkSZIkSZKm8v8Px3oXAeA17gAAAABJRU5ErkJggg==\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "# load the data\n", "data = np.loadtxt('DataSet_ML.txt')\n", "N = len(data)\n", "Nbins = 50\n", "\n", "# data\n", "X = data[:, :2]\n", "y = data[:, 2]\n", "\n", "# as signal and background\n", "sig = X[y == 1]\n", "bkg = X[y == 0]\n", "\n", "fig2, ax2 = plt.subplots(1, 3, figsize=(20, 8))\n", "ax2[0].hist(sig[:, 0], Nbins, histtype='step', label='sig', color='blue')\n", "ax2[0].hist(bkg[:, 0], Nbins, histtype='step', label='bkg', color='red')\n", "ax2[0].set(xlabel='Variable A', ylabel='Counts', title='Histogram of Variable A')\n", "ax2[0].legend()\n", "\n", "ax2[1].hist(sig[:, 1], Nbins, histtype='step', label='sig', color='blue')\n", "ax2[1].hist(bkg[:, 1], Nbins, histtype='step', label='bkg', color='red')\n", "ax2[1].set(xlabel='Variable B', ylabel='Counts', title='Histogram of Variable B')\n", "ax2[1].legend();" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# TASK:\n", "## 1) Think about how you think the above data looks in 2D before you continue.\n", "## 2) Draw on a piece of paper, what you think it looks like in 2D before you continue.\n", "When you have talked with your partner you should change `False` to `True` below:" ] }, { "cell_type": "code", "execution_count": 16, "metadata": {}, "outputs": [], "source": [ "if True:\n", "\n", " ax2[2].scatter(sig[:, 0], sig[:, 1], s=4, c='blue', label='sig', alpha=0.5)\n", " ax2[2].scatter(bkg[:, 0], bkg[:, 1], s=4, c='red', label='bkg', alpha=0.5)\n", " ax2[2].set(xlabel='Variable A', ylabel='Variable B', title='Scatterplot of Variable A and B')\n", " ax2[2].legend()\n", "\n", " fig2.tight_layout()\n", "\n", " fig2" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Interactive Machine Learning Part\n", "\n", "In this part we will further investigate different standard ML models. \n", "\n", "## Linear Fisher Discriminant:\n", "\n", "First we look at __[Linear Fisher Discriminant](https://scikit-learn.org/stable/modules/generated/sklearn.discriminant_analysis.LinearDiscriminantAnalysis.html)__. This is the same as we saw above so we will go through this example quite quickly (also because in 2D there are no interesting hyperparameters to change)." ] }, { "cell_type": "code", "execution_count": 33, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "# LDA\n", "clf_fisher = LDA(n_components=2) # initialise the LDA method\n", "clf_fisher.fit(X, y) # fit the data\n", "X_fisher = clf_fisher.transform(X) # transform the data\n", "#print(\"LDA coefficients\", clf_fisher.scalings_)\n", "\n", "# Extract the tranformed variables\n", "sig_fisher = X_fisher[y == 1]\n", "bkg_fisher = X_fisher[y == 0]\n", "\n", "# plot decision region of fisher\n", "fig_fisher, ax_fisher = plot_decision_regions(X, y, classifier=clf_fisher, title='Fisher')\n", "\n", "if plot_fisher_discriminant:\n", "\n", " # plot fisher discriminant \n", " fig_fisher2, ax_fisher2 = plt.subplots(figsize=(13.3, 6))\n", " ax_fisher2.hist(sig_fisher, Nbins, histtype='step', label='sig', color='blue')\n", " ax_fisher2.hist(bkg_fisher, Nbins, histtype='step', label='bkg', color='red')\n", " ax_fisher2.set(xlabel='Fisher Discriminant', ylabel='Counts', title='Fishers discriminant')\n", " ax_fisher2.legend()\n", " fig_fisher2.tight_layout()\n", " " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Decision Trees\n", "\n", "We load __[Decision Trees](https://scikit-learn.org/stable/modules/generated/sklearn.tree.DecisionTreeClassifier.html)__ from Scikit-learn (sklearn). At first try to increase the `max_depth` slider and see how that affects the plots. Does that make sense? For a given `max_depth`, e.g. `max_depth = 30`, change the `min_samples_leaf` and see how it simplifies (via _regularization_) the model. Think about when you'd want a simpler model. Finally, given a set of values for `max_depth` and `min_samples_leaf` switch between `criterion` being `gini` and `entropy`. Does this change much? " ] }, { "cell_type": "code", "execution_count": 37, "metadata": {}, "outputs": [ { "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "418cdbc45e0d44b187685a260ecc6f15", "version_major": 2, "version_minor": 0 }, "text/plain": [ "interactive(children=(Dropdown(description='criterion', options=('gini', 'entropy'), value='gini'), IntSlider(…" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "from sklearn.tree import DecisionTreeClassifier\n", "\n", "def animate_ML_estimator_DT(criterion, min_samples_leaf=1, max_depth=1):\n", " animate_ML_estimator_generator(DecisionTreeClassifier, 'Decision Tree', X, y, \n", " max_depth=max_depth, \n", " criterion=criterion,\n", " #splitter=splitter,\n", " #min_samples_split=min_samples_split, \n", " min_samples_leaf=min_samples_leaf)\n", "\n", "kwargs_DT = { 'max_depth': (1, 50), \n", " 'criterion': [\"gini\", \"entropy\"], \n", " #'splitter': [\"best\", \"random\"], \n", " #'min_samples_split': (2, 50),\n", " 'min_samples_leaf': (1, 50),\n", " } \n", "\n", "\n", "interactive_plot = interactive(animate_ML_estimator_DT, **kwargs_DT)\n", "interactive_plot" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Boosted Decision Trees (BDTs)\n", "\n", "For __[BDTs](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.AdaBoostClassifier.html#sklearn.ensemble.AdaBoostClassifier)__ try to slowly increase `n_estimators` and see how it affects the model. Does it make sense? Try also to increase the learning rate and see what that changes:\n" ] }, { "cell_type": "code", "execution_count": 38, "metadata": {}, "outputs": [ { "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "e3c3687bc13e44c9ac79caa79c42948a", "version_major": 2, "version_minor": 0 }, "text/plain": [ "interactive(children=(FloatSlider(value=1.0, description='learning_rate', max=2.0, min=0.01, step=0.01), IntSl…" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "from sklearn.ensemble import AdaBoostClassifier\n", "\n", "def animate_ML_estimator_BDT(learning_rate=1., n_estimators=1):\n", " animate_ML_estimator_generator(AdaBoostClassifier, 'Boosted Decision Trees', X, y, \n", " learning_rate=learning_rate, \n", " n_estimators=n_estimators,\n", " )\n", "\n", "kwargs_BDT = {'learning_rate': (0.01, 2, 0.01), \n", " 'n_estimators': (1, 100),\n", " } \n", "\n", "interactive_plot = interactive(animate_ML_estimator_BDT, **kwargs_BDT)\n", "interactive_plot" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## k-Nearest Neighbours\n", "\n", "For __[kNNs](https://scikit-learn.org/stable/modules/generated/sklearn.neighbors.KNeighborsClassifier.html)__ we only have the parameter n_neighbors to change:" ] }, { "cell_type": "code", "execution_count": 39, "metadata": {}, "outputs": [ { "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "0d4a7875520d41b69427acc8ce2f3457", "version_major": 2, "version_minor": 0 }, "text/plain": [ "interactive(children=(IntSlider(value=1, description='n_neighbors', max=50, min=1), Output()), _dom_classes=('…" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "from sklearn.neighbors import KNeighborsClassifier\n", "\n", "def animate_ML_estimator_KNN(n_neighbors=1):\n", " animate_ML_estimator_generator(KNeighborsClassifier, 'KNN', X, y, n_neighbors=n_neighbors)\n", "\n", "kwargs_KNN = {'n_neighbors': (1, 50), \n", " } \n", "\n", "interactive_plot = interactive(animate_ML_estimator_KNN, **kwargs_KNN)\n", "interactive_plot" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Support Vector Machine (SVM):\n", "\n", "__[SVms](https://scikit-learn.org/stable/modules/generated/sklearn.svm.SVC.html)__ are quite different ML models than the rest and we won't be going through it in much detail. However, see if you can make sense of the difference between the `RBF` kernel compared to the `poly` one. Also, how much does `C` matter for the different kernels?\n", "Notice that `degree` is only relevant for the `poly` kernel. " ] }, { "cell_type": "code", "execution_count": 40, "metadata": {}, "outputs": [ { "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "ad72cabadbf6487a972189ad1faabcf2", "version_major": 2, "version_minor": 0 }, "text/plain": [ "interactive(children=(Dropdown(description='kernel', index=1, options=('poly', 'rbf'), value='rbf'), FloatSlid…" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "from sklearn.svm import SVC\n", "\n", "def animate_ML_estimator_SVM(kernel='rbf', C=1, degree=2, ):\n", " animate_ML_estimator_generator(SVC, 'SVM', X, y, \n", " #gamma=gamma,\n", " gamma='scale',\n", " kernel=kernel, \n", " C=C,\n", " \n", " probability=True,\n", " degree=degree,\n", " )\n", "\n", "kwargs_SVM = {'C': (0.1, 10),\n", " 'kernel': ['poly', 'rbf'],\n", " 'degree': (1, 10),\n", " #'gamma': (0.1, 10),\n", " } \n", "\n", "interactive_plot = interactive(animate_ML_estimator_SVM, **kwargs_SVM)\n", "interactive_plot\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "***\n", "\n", " Machine Learning (ML) is a fascinating subject, which is very much in vouge these days.\n", " There are two classic usages for ML:\n", " \n", " - Classification (determine which category a case belongs to, i.e. ill or healthy)\n", " - Regression (determine a value, i.e. what was the energy of this electron)\n", "\n", " Python has packages - \"scikit-learn\" (among others) - that allows anybody to easily apply ML to smaller scale problems (i.e. below 10 GB). The following exercise is meant to illustrate a classification problem, and to wet your appetite for more...\n", "\n", "# Questions:\n", "\n", "1. Consider the data, and make sure that you by eye can see, how an algorithm should decide between the two categories. Also see, if you can guess how well the Fisher performs. What is the error rate of type1 and type2 roughly?\n", "\n", "2. Now consider the various ML algorithms. Can you (again by eye) tell, if it is doing \"just well\" or \"very well\"? I.e. can you rank the methods by eye? What you have tried this, compare to the ROC curves and see how well you did.\n", "\n", "3. What we are currently plotting in the ROCcurve is the ROC-curve for the training data. Try to split up the data into a training and a test set and add the ROC-curve of the test set to the `plot_decision_regions` function. \n", "\n", "4. Try to put all the ROC curves into one final plot, which shows how well the different methods perform." ] } ], "metadata": { "executable": "/usr/bin/env python", "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": "3.6.7" }, "main_language": "python" }, "nbformat": 4, "nbformat_minor": 2 }