cap2.9 - Neural Networks (Deep Learning).ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Neural Networks (Deep Learning)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Usually Deep learning algo. are very tailored to a specific use case. Here we'll discuss only Multilayers Perceptrons (MLPs)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1.0 The neural network model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "MLPs can be viewed as generalizations of linear models."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In linear regression we assume that:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "y_hat = w[0]*x[0] + ... + w[p]*x[p] + b"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-22T01:59:58.928571Z",
     "start_time": "2020-12-22T01:59:56.368563Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<?xml version=\"1.0\" encoding=\"UTF-8\" standalone=\"no\"?>\n",
       "<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\n",
       " \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\n",
       "<!-- Generated by graphviz version 2.44.1 (20200629.0846)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"174pt\" height=\"261pt\"\n",
       " viewBox=\"0.00 0.00 174.00 261.00\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       "<g id=\"graph0\" class=\"graph\" transform=\"scale(1 1) rotate(0) translate(4 257)\">\n",
       "<polygon fill=\"white\" stroke=\"transparent\" points=\"-4,4 -4,-257 170,-257 170,4 -4,4\"/>\n",
       "<g id=\"clust1\" class=\"cluster\">\n",
       "<title>cluster_0</title>\n",
       "<polygon fill=\"none\" stroke=\"white\" points=\"8,-8 8,-245 60,-245 60,-8 8,-8\"/>\n",
       "<text text-anchor=\"middle\" x=\"34\" y=\"-229.8\" font-family=\"Times,serif\" font-size=\"14.00\">inputs</text>\n",
       "</g>\n",
       "<g id=\"clust2\" class=\"cluster\">\n",
       "<title>cluster_2</title>\n",
       "<polygon fill=\"none\" stroke=\"white\" points=\"106,-89 106,-164 158,-164 158,-89 106,-89\"/>\n",
       "<text text-anchor=\"middle\" x=\"132\" y=\"-148.8\" font-family=\"Times,serif\" font-size=\"14.00\">output</text>\n",
       "</g>\n",
       "<!-- x[0] -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>x[0]</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"34\" cy=\"-196\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"34\" y=\"-192.3\" font-family=\"Times,serif\" font-size=\"14.00\">x[0]</text>\n",
       "</g>\n",
       "<!-- y -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>y</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"132\" cy=\"-115\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"132\" y=\"-111.3\" font-family=\"Times,serif\" font-size=\"14.00\">y</text>\n",
       "</g>\n",
       "<!-- x[0]&#45;&gt;y -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>x[0]&#45;&gt;y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M48.35,-184.74C64.33,-171.26 91.27,-148.52 110.03,-132.69\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"112.38,-135.29 117.76,-126.17 107.86,-129.94 112.38,-135.29\"/>\n",
       "<text text-anchor=\"middle\" x=\"83\" y=\"-171.8\" font-family=\"Times,serif\" font-size=\"14.00\">w[0]</text>\n",
       "</g>\n",
       "<!-- x[1] -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>x[1]</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"34\" cy=\"-142\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"34\" y=\"-138.3\" font-family=\"Times,serif\" font-size=\"14.00\">x[1]</text>\n",
       "</g>\n",
       "<!-- x[1]&#45;&gt;y -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>x[1]&#45;&gt;y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M51.72,-137.3C66.31,-133.19 87.74,-127.17 104.63,-122.42\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"105.79,-125.73 114.47,-119.65 103.9,-118.99 105.79,-125.73\"/>\n",
       "<text text-anchor=\"middle\" x=\"83\" y=\"-134.8\" font-family=\"Times,serif\" font-size=\"14.00\">w[1]</text>\n",
       "</g>\n",
       "<!-- x[2] -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>x[2]</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"34\" cy=\"-88\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"34\" y=\"-84.3\" font-family=\"Times,serif\" font-size=\"14.00\">x[2]</text>\n",
       "</g>\n",
       "<!-- x[2]&#45;&gt;y -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>x[2]&#45;&gt;y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M51.72,-92.7C66.31,-96.81 87.74,-102.83 104.63,-107.58\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"103.9,-111.01 114.47,-110.35 105.79,-104.27 103.9,-111.01\"/>\n",
       "<text text-anchor=\"middle\" x=\"83\" y=\"-107.8\" font-family=\"Times,serif\" font-size=\"14.00\">w[2]</text>\n",
       "</g>\n",
       "<!-- x[3] -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>x[3]</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"34\" cy=\"-34\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"34\" y=\"-30.3\" font-family=\"Times,serif\" font-size=\"14.00\">x[3]</text>\n",
       "</g>\n",
       "<!-- x[3]&#45;&gt;y -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>x[3]&#45;&gt;y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M48.35,-45.26C64.33,-58.74 91.27,-81.48 110.03,-97.31\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"107.86,-100.06 117.76,-103.83 112.38,-94.71 107.86,-100.06\"/>\n",
       "<text text-anchor=\"middle\" x=\"83\" y=\"-82.8\" font-family=\"Times,serif\" font-size=\"14.00\">w[3]</text>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.dot.Digraph at 0x7ff7504e7850>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import mglearn\n",
    "display(mglearn.plots.plot_logistic_regression_graph())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-22T02:02:32.094095Z",
     "start_time": "2020-12-22T02:02:31.828006Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<?xml version=\"1.0\" encoding=\"UTF-8\" standalone=\"no\"?>\n",
       "<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\n",
       " \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\n",
       "<!-- Generated by graphviz version 2.44.1 (20200629.0846)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"252pt\" height=\"261pt\"\n",
       " viewBox=\"0.00 0.00 252.00 261.00\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       "<g id=\"graph0\" class=\"graph\" transform=\"scale(1 1) rotate(0) translate(4 257)\">\n",
       "<polygon fill=\"white\" stroke=\"transparent\" points=\"-4,4 -4,-257 248,-257 248,4 -4,4\"/>\n",
       "<g id=\"clust1\" class=\"cluster\">\n",
       "<title>cluster_0</title>\n",
       "<polygon fill=\"none\" stroke=\"white\" points=\"8,-8 8,-245 60,-245 60,-8 8,-8\"/>\n",
       "<text text-anchor=\"middle\" x=\"34\" y=\"-229.8\" font-family=\"Times,serif\" font-size=\"14.00\">inputs</text>\n",
       "</g>\n",
       "<g id=\"clust2\" class=\"cluster\">\n",
       "<title>cluster_1</title>\n",
       "<polygon fill=\"none\" stroke=\"white\" points=\"80,-35 80,-218 164,-218 164,-35 80,-35\"/>\n",
       "<text text-anchor=\"middle\" x=\"122\" y=\"-202.8\" font-family=\"Times,serif\" font-size=\"14.00\">hidden layer</text>\n",
       "</g>\n",
       "<g id=\"clust3\" class=\"cluster\">\n",
       "<title>cluster_2</title>\n",
       "<polygon fill=\"none\" stroke=\"white\" points=\"184,-89 184,-164 236,-164 236,-89 184,-89\"/>\n",
       "<text text-anchor=\"middle\" x=\"210\" y=\"-148.8\" font-family=\"Times,serif\" font-size=\"14.00\">output</text>\n",
       "</g>\n",
       "<!-- x[0] -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>x[0]</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"34\" cy=\"-196\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"34\" y=\"-192.3\" font-family=\"Times,serif\" font-size=\"14.00\">x[0]</text>\n",
       "</g>\n",
       "<!-- h0 -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>h0</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"122\" cy=\"-61\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"122\" y=\"-57.3\" font-family=\"Times,serif\" font-size=\"14.00\">h[0]</text>\n",
       "</g>\n",
       "<!-- x[0]&#45;&gt;h0 -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>x[0]&#45;&gt;h0</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M44.53,-181.04C59.38,-157.72 88.18,-112.52 105.88,-84.73\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"109,-86.34 111.43,-76.03 103.1,-82.58 109,-86.34\"/>\n",
       "</g>\n",
       "<!-- h1 -->\n",
       "<g id=\"node6\" class=\"node\">\n",
       "<title>h1</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"122\" cy=\"-169\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"122\" y=\"-165.3\" font-family=\"Times,serif\" font-size=\"14.00\">h[1]</text>\n",
       "</g>\n",
       "<!-- x[0]&#45;&gt;h1 -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>x[0]&#45;&gt;h1</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M51.58,-190.79C63.84,-186.95 80.76,-181.63 94.84,-177.21\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"95.98,-180.52 104.47,-174.19 93.88,-173.85 95.98,-180.52\"/>\n",
       "</g>\n",
       "<!-- h2 -->\n",
       "<g id=\"node7\" class=\"node\">\n",
       "<title>h2</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"122\" cy=\"-115\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"122\" y=\"-111.3\" font-family=\"Times,serif\" font-size=\"14.00\">h[2]</text>\n",
       "</g>\n",
       "<!-- x[0]&#45;&gt;h2 -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>x[0]&#45;&gt;h2</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M47.69,-184.05C61.66,-170.89 84.32,-149.54 100.76,-134.07\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"103.38,-136.41 108.26,-127 98.58,-131.31 103.38,-136.41\"/>\n",
       "</g>\n",
       "<!-- x[1] -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>x[1]</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"34\" cy=\"-142\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"34\" y=\"-138.3\" font-family=\"Times,serif\" font-size=\"14.00\">x[1]</text>\n",
       "</g>\n",
       "<!-- x[1]&#45;&gt;h0 -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>x[1]&#45;&gt;h0</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M47.69,-130.05C61.66,-116.89 84.32,-95.54 100.76,-80.07\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"103.38,-82.41 108.26,-73 98.58,-77.31 103.38,-82.41\"/>\n",
       "</g>\n",
       "<!-- x[1]&#45;&gt;h1 -->\n",
       "<g id=\"edge5\" class=\"edge\">\n",
       "<title>x[1]&#45;&gt;h1</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M51.58,-147.21C63.84,-151.05 80.76,-156.37 94.84,-160.79\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"93.88,-164.15 104.47,-163.81 95.98,-157.48 93.88,-164.15\"/>\n",
       "</g>\n",
       "<!-- x[1]&#45;&gt;h2 -->\n",
       "<g id=\"edge6\" class=\"edge\">\n",
       "<title>x[1]&#45;&gt;h2</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M51.58,-136.79C63.84,-132.95 80.76,-127.63 94.84,-123.21\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"95.98,-126.52 104.47,-120.19 93.88,-119.85 95.98,-126.52\"/>\n",
       "</g>\n",
       "<!-- x[2] -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>x[2]</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"34\" cy=\"-88\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"34\" y=\"-84.3\" font-family=\"Times,serif\" font-size=\"14.00\">x[2]</text>\n",
       "</g>\n",
       "<!-- x[2]&#45;&gt;h0 -->\n",
       "<g id=\"edge7\" class=\"edge\">\n",
       "<title>x[2]&#45;&gt;h0</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M51.58,-82.79C63.84,-78.95 80.76,-73.63 94.84,-69.21\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"95.98,-72.52 104.47,-66.19 93.88,-65.85 95.98,-72.52\"/>\n",
       "</g>\n",
       "<!-- x[2]&#45;&gt;h1 -->\n",
       "<g id=\"edge8\" class=\"edge\">\n",
       "<title>x[2]&#45;&gt;h1</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M47.69,-99.95C61.66,-113.11 84.32,-134.46 100.76,-149.93\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"98.58,-152.69 108.26,-157 103.38,-147.59 98.58,-152.69\"/>\n",
       "</g>\n",
       "<!-- x[2]&#45;&gt;h2 -->\n",
       "<g id=\"edge9\" class=\"edge\">\n",
       "<title>x[2]&#45;&gt;h2</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M51.58,-93.21C63.84,-97.05 80.76,-102.37 94.84,-106.79\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"93.88,-110.15 104.47,-109.81 95.98,-103.48 93.88,-110.15\"/>\n",
       "</g>\n",
       "<!-- x[3] -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>x[3]</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"34\" cy=\"-34\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"34\" y=\"-30.3\" font-family=\"Times,serif\" font-size=\"14.00\">x[3]</text>\n",
       "</g>\n",
       "<!-- x[3]&#45;&gt;h0 -->\n",
       "<g id=\"edge10\" class=\"edge\">\n",
       "<title>x[3]&#45;&gt;h0</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M51.58,-39.21C63.84,-43.05 80.76,-48.37 94.84,-52.79\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"93.88,-56.15 104.47,-55.81 95.98,-49.48 93.88,-56.15\"/>\n",
       "</g>\n",
       "<!-- x[3]&#45;&gt;h1 -->\n",
       "<g id=\"edge11\" class=\"edge\">\n",
       "<title>x[3]&#45;&gt;h1</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M44.53,-48.96C59.38,-72.28 88.18,-117.48 105.88,-145.27\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"103.1,-147.42 111.43,-153.97 109,-143.66 103.1,-147.42\"/>\n",
       "</g>\n",
       "<!-- x[3]&#45;&gt;h2 -->\n",
       "<g id=\"edge12\" class=\"edge\">\n",
       "<title>x[3]&#45;&gt;h2</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M47.69,-45.95C61.66,-59.11 84.32,-80.46 100.76,-95.93\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"98.58,-98.69 108.26,-103 103.38,-93.59 98.58,-98.69\"/>\n",
       "</g>\n",
       "<!-- y -->\n",
       "<g id=\"node8\" class=\"node\">\n",
       "<title>y</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"210\" cy=\"-115\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"210\" y=\"-111.3\" font-family=\"Times,serif\" font-size=\"14.00\">y</text>\n",
       "</g>\n",
       "<!-- h0&#45;&gt;y -->\n",
       "<g id=\"edge13\" class=\"edge\">\n",
       "<title>h0&#45;&gt;y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M137.59,-70.16C150.68,-78.38 170.15,-90.6 185.45,-100.21\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"184.01,-103.44 194.34,-105.79 187.73,-97.51 184.01,-103.44\"/>\n",
       "</g>\n",
       "<!-- h1&#45;&gt;y -->\n",
       "<g id=\"edge14\" class=\"edge\">\n",
       "<title>h1&#45;&gt;y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M137.59,-159.84C150.68,-151.62 170.15,-139.4 185.45,-129.79\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"187.73,-132.49 194.34,-124.21 184.01,-126.56 187.73,-132.49\"/>\n",
       "</g>\n",
       "<!-- h2&#45;&gt;y -->\n",
       "<g id=\"edge15\" class=\"edge\">\n",
       "<title>h2&#45;&gt;y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M140.4,-115C152.25,-115 168.18,-115 181.71,-115\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"181.92,-118.5 191.92,-115 181.92,-111.5 181.92,-118.5\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.dot.Digraph at 0x7ff754f3d6a0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(mglearn.plots.plot_single_hidden_layer_graph())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "h[0] = np.tanh(w[0, 0] * x[0] + w[1, 0] * x[1] + w[2, 0] * x[2] + w[3, 0] * x[3])\n",
    "\n",
    "h[1] = np.tanh(w[0, 0] * x[0] + w[1, 0] * x[1] + w[2, 0] * x[2] + w[3, 0] * x[3])\n",
    "\n",
    "h[2] = np.tanh(w[0, 0] * x[0] + w[1, 0] * x[1] + w[2, 0] * x[2] + w[3, 0] * x[3])\n",
    "\n",
    "ŷ = v[0] * h[0] + v[1] * h[1] + v[2] * h[2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-22T02:21:09.026085Z",
     "start_time": "2020-12-22T02:21:08.809112Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "line = np.linspace(-3,3,100)\n",
    "plt.plot(line, np.tanh(line), label='tanh') # tangent activation function\n",
    "plt.plot(line, np.maximum(line, 0), label = 'relu') #rectified linear activation function\n",
    "plt.legend(loc='best')\n",
    "plt.xlabel('x')\n",
    "plt.ylabel('relu(x), tahn(x)');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-22T02:33:55.865883Z",
     "start_time": "2020-12-22T02:33:55.531101Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<?xml version=\"1.0\" encoding=\"UTF-8\" standalone=\"no\"?>\n",
       "<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\n",
       " \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\n",
       "<!-- Generated by graphviz version 2.44.1 (20200629.0846)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"378pt\" height=\"261pt\"\n",
       " viewBox=\"0.00 0.00 378.00 261.00\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       "<g id=\"graph0\" class=\"graph\" transform=\"scale(1 1) rotate(0) translate(4 257)\">\n",
       "<polygon fill=\"white\" stroke=\"transparent\" points=\"-4,4 -4,-257 374,-257 374,4 -4,4\"/>\n",
       "<g id=\"clust1\" class=\"cluster\">\n",
       "<title>cluster_0</title>\n",
       "<polygon fill=\"none\" stroke=\"white\" points=\"8,-8 8,-245 60,-245 60,-8 8,-8\"/>\n",
       "<text text-anchor=\"middle\" x=\"34\" y=\"-229.8\" font-family=\"Times,serif\" font-size=\"14.00\">inputs</text>\n",
       "</g>\n",
       "<g id=\"clust2\" class=\"cluster\">\n",
       "<title>cluster_1</title>\n",
       "<polygon fill=\"none\" stroke=\"white\" points=\"80,-35 80,-218 175,-218 175,-35 80,-35\"/>\n",
       "<text text-anchor=\"middle\" x=\"127.5\" y=\"-202.8\" font-family=\"Times,serif\" font-size=\"14.00\">hidden layer 1</text>\n",
       "</g>\n",
       "<g id=\"clust3\" class=\"cluster\">\n",
       "<title>cluster_2</title>\n",
       "<polygon fill=\"none\" stroke=\"white\" points=\"195,-35 195,-218 290,-218 290,-35 195,-35\"/>\n",
       "<text text-anchor=\"middle\" x=\"242.5\" y=\"-202.8\" font-family=\"Times,serif\" font-size=\"14.00\">hidden layer 2</text>\n",
       "</g>\n",
       "<g id=\"clust4\" class=\"cluster\">\n",
       "<title>cluster_3</title>\n",
       "<polygon fill=\"none\" stroke=\"white\" points=\"310,-89 310,-164 362,-164 362,-89 310,-89\"/>\n",
       "<text text-anchor=\"middle\" x=\"336\" y=\"-148.8\" font-family=\"Times,serif\" font-size=\"14.00\">output</text>\n",
       "</g>\n",
       "<!-- x[0] -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>x[0]</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"34\" cy=\"-196\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"34\" y=\"-192.3\" font-family=\"Times,serif\" font-size=\"14.00\">x[0]</text>\n",
       "</g>\n",
       "<!-- h1[0] -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>h1[0]</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"127\" cy=\"-61\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"127\" y=\"-57.3\" font-family=\"Times,serif\" font-size=\"14.00\">h1[0]</text>\n",
       "</g>\n",
       "<!-- x[0]&#45;&gt;h1[0] -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>x[0]&#45;&gt;h1[0]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M44.74,-181.55C60.46,-158.23 91.57,-112.08 110.41,-84.13\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"113.37,-86 116.06,-75.75 107.56,-82.09 113.37,-86\"/>\n",
       "</g>\n",
       "<!-- h1[1] -->\n",
       "<g id=\"node6\" class=\"node\">\n",
       "<title>h1[1]</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"127\" cy=\"-169\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"127\" y=\"-165.3\" font-family=\"Times,serif\" font-size=\"14.00\">h1[1]</text>\n",
       "</g>\n",
       "<!-- x[0]&#45;&gt;h1[1] -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>x[0]&#45;&gt;h1[1]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M51.69,-191.05C65.03,-187.09 83.99,-181.47 99.41,-176.89\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"100.82,-180.12 109.42,-173.92 98.83,-173.41 100.82,-180.12\"/>\n",
       "</g>\n",
       "<!-- h1[2] -->\n",
       "<g id=\"node7\" class=\"node\">\n",
       "<title>h1[2]</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"127\" cy=\"-115\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"127\" y=\"-111.3\" font-family=\"Times,serif\" font-size=\"14.00\">h1[2]</text>\n",
       "</g>\n",
       "<!-- x[0]&#45;&gt;h1[2] -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>x[0]&#45;&gt;h1[2]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M48.04,-184.39C63.04,-171.04 87.86,-148.94 105.47,-133.28\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"107.84,-135.85 112.99,-126.58 103.19,-130.62 107.84,-135.85\"/>\n",
       "</g>\n",
       "<!-- x[1] -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>x[1]</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"34\" cy=\"-142\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"34\" y=\"-138.3\" font-family=\"Times,serif\" font-size=\"14.00\">x[1]</text>\n",
       "</g>\n",
       "<!-- x[1]&#45;&gt;h1[0] -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>x[1]&#45;&gt;h1[0]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M48.04,-130.39C63.04,-117.04 87.86,-94.94 105.47,-79.28\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"107.84,-81.85 112.99,-72.58 103.19,-76.62 107.84,-81.85\"/>\n",
       "</g>\n",
       "<!-- x[1]&#45;&gt;h1[1] -->\n",
       "<g id=\"edge5\" class=\"edge\">\n",
       "<title>x[1]&#45;&gt;h1[1]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M51.69,-146.95C65.03,-150.91 83.99,-156.53 99.41,-161.11\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"98.83,-164.59 109.42,-164.08 100.82,-157.88 98.83,-164.59\"/>\n",
       "</g>\n",
       "<!-- x[1]&#45;&gt;h1[2] -->\n",
       "<g id=\"edge6\" class=\"edge\">\n",
       "<title>x[1]&#45;&gt;h1[2]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M51.69,-137.05C65.03,-133.09 83.99,-127.47 99.41,-122.89\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"100.82,-126.12 109.42,-119.92 98.83,-119.41 100.82,-126.12\"/>\n",
       "</g>\n",
       "<!-- x[2] -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>x[2]</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"34\" cy=\"-88\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"34\" y=\"-84.3\" font-family=\"Times,serif\" font-size=\"14.00\">x[2]</text>\n",
       "</g>\n",
       "<!-- x[2]&#45;&gt;h1[0] -->\n",
       "<g id=\"edge7\" class=\"edge\">\n",
       "<title>x[2]&#45;&gt;h1[0]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M51.69,-83.05C65.03,-79.09 83.99,-73.47 99.41,-68.89\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"100.82,-72.12 109.42,-65.92 98.83,-65.41 100.82,-72.12\"/>\n",
       "</g>\n",
       "<!-- x[2]&#45;&gt;h1[1] -->\n",
       "<g id=\"edge8\" class=\"edge\">\n",
       "<title>x[2]&#45;&gt;h1[1]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M48.04,-99.61C63.04,-112.96 87.86,-135.06 105.47,-150.72\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"103.19,-153.38 112.99,-157.42 107.84,-148.15 103.19,-153.38\"/>\n",
       "</g>\n",
       "<!-- x[2]&#45;&gt;h1[2] -->\n",
       "<g id=\"edge9\" class=\"edge\">\n",
       "<title>x[2]&#45;&gt;h1[2]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M51.69,-92.95C65.03,-96.91 83.99,-102.53 99.41,-107.11\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"98.83,-110.59 109.42,-110.08 100.82,-103.88 98.83,-110.59\"/>\n",
       "</g>\n",
       "<!-- x[3] -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>x[3]</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"34\" cy=\"-34\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"34\" y=\"-30.3\" font-family=\"Times,serif\" font-size=\"14.00\">x[3]</text>\n",
       "</g>\n",
       "<!-- x[3]&#45;&gt;h1[0] -->\n",
       "<g id=\"edge10\" class=\"edge\">\n",
       "<title>x[3]&#45;&gt;h1[0]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M51.69,-38.95C65.03,-42.91 83.99,-48.53 99.41,-53.11\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"98.83,-56.59 109.42,-56.08 100.82,-49.88 98.83,-56.59\"/>\n",
       "</g>\n",
       "<!-- x[3]&#45;&gt;h1[1] -->\n",
       "<g id=\"edge11\" class=\"edge\">\n",
       "<title>x[3]&#45;&gt;h1[1]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M44.74,-48.45C60.46,-71.77 91.57,-117.92 110.41,-145.87\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"107.56,-147.91 116.06,-154.25 113.37,-144 107.56,-147.91\"/>\n",
       "</g>\n",
       "<!-- x[3]&#45;&gt;h1[2] -->\n",
       "<g id=\"edge12\" class=\"edge\">\n",
       "<title>x[3]&#45;&gt;h1[2]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M48.04,-45.61C63.04,-58.96 87.86,-81.06 105.47,-96.72\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"103.19,-99.38 112.99,-103.42 107.84,-94.15 103.19,-99.38\"/>\n",
       "</g>\n",
       "<!-- h2[0] -->\n",
       "<g id=\"node8\" class=\"node\">\n",
       "<title>h2[0]</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"242\" cy=\"-61\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"242\" y=\"-57.3\" font-family=\"Times,serif\" font-size=\"14.00\">h2[0]</text>\n",
       "</g>\n",
       "<!-- h1[0]&#45;&gt;h2[0] -->\n",
       "<g id=\"edge13\" class=\"edge\">\n",
       "<title>h1[0]&#45;&gt;h2[0]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M145.16,-61C163.33,-61 192.28,-61 213.61,-61\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"213.68,-64.5 223.68,-61 213.68,-57.5 213.68,-64.5\"/>\n",
       "</g>\n",
       "<!-- h2[1] -->\n",
       "<g id=\"node9\" class=\"node\">\n",
       "<title>h2[1]</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"242\" cy=\"-169\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"242\" y=\"-165.3\" font-family=\"Times,serif\" font-size=\"14.00\">h2[1]</text>\n",
       "</g>\n",
       "<!-- h1[0]&#45;&gt;h2[1] -->\n",
       "<g id=\"edge14\" class=\"edge\">\n",
       "<title>h1[0]&#45;&gt;h2[1]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M140.52,-72.97C160.04,-91.62 197.87,-127.78 221.14,-150.01\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"218.77,-152.6 228.42,-156.98 223.61,-147.54 218.77,-152.6\"/>\n",
       "</g>\n",
       "<!-- h2[2] -->\n",
       "<g id=\"node10\" class=\"node\">\n",
       "<title>h2[2]</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"242\" cy=\"-115\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"242\" y=\"-111.3\" font-family=\"Times,serif\" font-size=\"14.00\">h2[2]</text>\n",
       "</g>\n",
       "<!-- h1[0]&#45;&gt;h2[2] -->\n",
       "<g id=\"edge15\" class=\"edge\">\n",
       "<title>h1[0]&#45;&gt;h2[2]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M143.71,-68.51C162.52,-77.5 194.23,-92.65 216.32,-103.21\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"214.89,-106.4 225.42,-107.56 217.91,-100.09 214.89,-106.4\"/>\n",
       "</g>\n",
       "<!-- h1[1]&#45;&gt;h2[0] -->\n",
       "<g id=\"edge16\" class=\"edge\">\n",
       "<title>h1[1]&#45;&gt;h2[0]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M140.52,-157.03C160.04,-138.38 197.87,-102.22 221.14,-79.99\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"223.61,-82.46 228.42,-73.02 218.77,-77.4 223.61,-82.46\"/>\n",
       "</g>\n",
       "<!-- h1[1]&#45;&gt;h2[1] -->\n",
       "<g id=\"edge17\" class=\"edge\">\n",
       "<title>h1[1]&#45;&gt;h2[1]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M145.16,-169C163.33,-169 192.28,-169 213.61,-169\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"213.68,-172.5 223.68,-169 213.68,-165.5 213.68,-172.5\"/>\n",
       "</g>\n",
       "<!-- h1[1]&#45;&gt;h2[2] -->\n",
       "<g id=\"edge18\" class=\"edge\">\n",
       "<title>h1[1]&#45;&gt;h2[2]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M143.71,-161.49C162.52,-152.5 194.23,-137.35 216.32,-126.79\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"217.91,-129.91 225.42,-122.44 214.89,-123.6 217.91,-129.91\"/>\n",
       "</g>\n",
       "<!-- h1[2]&#45;&gt;h2[0] -->\n",
       "<g id=\"edge19\" class=\"edge\">\n",
       "<title>h1[2]&#45;&gt;h2[0]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M143.71,-107.49C162.52,-98.5 194.23,-83.35 216.32,-72.79\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"217.91,-75.91 225.42,-68.44 214.89,-69.6 217.91,-75.91\"/>\n",
       "</g>\n",
       "<!-- h1[2]&#45;&gt;h2[1] -->\n",
       "<g id=\"edge20\" class=\"edge\">\n",
       "<title>h1[2]&#45;&gt;h2[1]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M143.71,-122.51C162.52,-131.5 194.23,-146.65 216.32,-157.21\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"214.89,-160.4 225.42,-161.56 217.91,-154.09 214.89,-160.4\"/>\n",
       "</g>\n",
       "<!-- h1[2]&#45;&gt;h2[2] -->\n",
       "<g id=\"edge21\" class=\"edge\">\n",
       "<title>h1[2]&#45;&gt;h2[2]</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M145.16,-115C163.33,-115 192.28,-115 213.61,-115\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"213.68,-118.5 223.68,-115 213.68,-111.5 213.68,-118.5\"/>\n",
       "</g>\n",
       "<!-- y -->\n",
       "<g id=\"node11\" class=\"node\">\n",
       "<title>y</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"336\" cy=\"-115\" rx=\"18\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"336\" y=\"-111.3\" font-family=\"Times,serif\" font-size=\"14.00\">y</text>\n",
       "</g>\n",
       "<!-- h2[0]&#45;&gt;y -->\n",
       "<g id=\"edge22\" class=\"edge\">\n",
       "<title>h2[0]&#45;&gt;y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M257.78,-69.68C272.18,-78.13 294.31,-91.12 311.19,-101.03\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"309.55,-104.12 319.95,-106.17 313.1,-98.09 309.55,-104.12\"/>\n",
       "</g>\n",
       "<!-- h2[1]&#45;&gt;y -->\n",
       "<g id=\"edge23\" class=\"edge\">\n",
       "<title>h2[1]&#45;&gt;y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M257.78,-160.32C272.18,-151.87 294.31,-138.88 311.19,-128.97\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"313.1,-131.91 319.95,-123.83 309.55,-125.88 313.1,-131.91\"/>\n",
       "</g>\n",
       "<!-- h2[2]&#45;&gt;y -->\n",
       "<g id=\"edge24\" class=\"edge\">\n",
       "<title>h2[2]&#45;&gt;y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M260.3,-115C273.54,-115 292.05,-115 307.34,-115\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"307.77,-118.5 317.77,-115 307.77,-111.5 307.77,-118.5\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.dot.Digraph at 0x7ff75539a580>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mglearn.plots.plot_two_hidden_layer_graph()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Large neural networks made up of many layers of computation is what inspired the name deep learning."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-22T02:36:47.851650Z",
     "start_time": "2020-12-22T02:36:47.843049Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Help on class MLPClassifier in module sklearn.neural_network._multilayer_perceptron:\n",
      "\n",
      "class MLPClassifier(sklearn.base.ClassifierMixin, BaseMultilayerPerceptron)\n",
      " |  MLPClassifier(hidden_layer_sizes=(100,), activation='relu', *, solver='adam', alpha=0.0001, batch_size='auto', learning_rate='constant', learning_rate_init=0.001, power_t=0.5, max_iter=200, shuffle=True, random_state=None, tol=0.0001, verbose=False, warm_start=False, momentum=0.9, nesterovs_momentum=True, early_stopping=False, validation_fraction=0.1, beta_1=0.9, beta_2=0.999, epsilon=1e-08, n_iter_no_change=10, max_fun=15000)\n",
      " |  \n",
      " |  Multi-layer Perceptron classifier.\n",
      " |  \n",
      " |  This model optimizes the log-loss function using LBFGS or stochastic\n",
      " |  gradient descent.\n",
      " |  \n",
      " |  .. versionadded:: 0.18\n",
      " |  \n",
      " |  Parameters\n",
      " |  ----------\n",
      " |  hidden_layer_sizes : tuple, length = n_layers - 2, default=(100,)\n",
      " |      The ith element represents the number of neurons in the ith\n",
      " |      hidden layer.\n",
      " |  \n",
      " |  activation : {'identity', 'logistic', 'tanh', 'relu'}, default='relu'\n",
      " |      Activation function for the hidden layer.\n",
      " |  \n",
      " |      - 'identity', no-op activation, useful to implement linear bottleneck,\n",
      " |        returns f(x) = x\n",
      " |  \n",
      " |      - 'logistic', the logistic sigmoid function,\n",
      " |        returns f(x) = 1 / (1 + exp(-x)).\n",
      " |  \n",
      " |      - 'tanh', the hyperbolic tan function,\n",
      " |        returns f(x) = tanh(x).\n",
      " |  \n",
      " |      - 'relu', the rectified linear unit function,\n",
      " |        returns f(x) = max(0, x)\n",
      " |  \n",
      " |  solver : {'lbfgs', 'sgd', 'adam'}, default='adam'\n",
      " |      The solver for weight optimization.\n",
      " |  \n",
      " |      - 'lbfgs' is an optimizer in the family of quasi-Newton methods.\n",
      " |  \n",
      " |      - 'sgd' refers to stochastic gradient descent.\n",
      " |  \n",
      " |      - 'adam' refers to a stochastic gradient-based optimizer proposed\n",
      " |        by Kingma, Diederik, and Jimmy Ba\n",
      " |  \n",
      " |      Note: The default solver 'adam' works pretty well on relatively\n",
      " |      large datasets (with thousands of training samples or more) in terms of\n",
      " |      both training time and validation score.\n",
      " |      For small datasets, however, 'lbfgs' can converge faster and perform\n",
      " |      better.\n",
      " |  \n",
      " |  alpha : float, default=0.0001\n",
      " |      L2 penalty (regularization term) parameter.\n",
      " |  \n",
      " |  batch_size : int, default='auto'\n",
      " |      Size of minibatches for stochastic optimizers.\n",
      " |      If the solver is 'lbfgs', the classifier will not use minibatch.\n",
      " |      When set to \"auto\", `batch_size=min(200, n_samples)`\n",
      " |  \n",
      " |  learning_rate : {'constant', 'invscaling', 'adaptive'}, default='constant'\n",
      " |      Learning rate schedule for weight updates.\n",
      " |  \n",
      " |      - 'constant' is a constant learning rate given by\n",
      " |        'learning_rate_init'.\n",
      " |  \n",
      " |      - 'invscaling' gradually decreases the learning rate at each\n",
      " |        time step 't' using an inverse scaling exponent of 'power_t'.\n",
      " |        effective_learning_rate = learning_rate_init / pow(t, power_t)\n",
      " |  \n",
      " |      - 'adaptive' keeps the learning rate constant to\n",
      " |        'learning_rate_init' as long as training loss keeps decreasing.\n",
      " |        Each time two consecutive epochs fail to decrease training loss by at\n",
      " |        least tol, or fail to increase validation score by at least tol if\n",
      " |        'early_stopping' is on, the current learning rate is divided by 5.\n",
      " |  \n",
      " |      Only used when ``solver='sgd'``.\n",
      " |  \n",
      " |  learning_rate_init : double, default=0.001\n",
      " |      The initial learning rate used. It controls the step-size\n",
      " |      in updating the weights. Only used when solver='sgd' or 'adam'.\n",
      " |  \n",
      " |  power_t : double, default=0.5\n",
      " |      The exponent for inverse scaling learning rate.\n",
      " |      It is used in updating effective learning rate when the learning_rate\n",
      " |      is set to 'invscaling'. Only used when solver='sgd'.\n",
      " |  \n",
      " |  max_iter : int, default=200\n",
      " |      Maximum number of iterations. The solver iterates until convergence\n",
      " |      (determined by 'tol') or this number of iterations. For stochastic\n",
      " |      solvers ('sgd', 'adam'), note that this determines the number of epochs\n",
      " |      (how many times each data point will be used), not the number of\n",
      " |      gradient steps.\n",
      " |  \n",
      " |  shuffle : bool, default=True\n",
      " |      Whether to shuffle samples in each iteration. Only used when\n",
      " |      solver='sgd' or 'adam'.\n",
      " |  \n",
      " |  random_state : int, RandomState instance, default=None\n",
      " |      Determines random number generation for weights and bias\n",
      " |      initialization, train-test split if early stopping is used, and batch\n",
      " |      sampling when solver='sgd' or 'adam'.\n",
      " |      Pass an int for reproducible results across multiple function calls.\n",
      " |      See :term:`Glossary <random_state>`.\n",
      " |  \n",
      " |  tol : float, default=1e-4\n",
      " |      Tolerance for the optimization. When the loss or score is not improving\n",
      " |      by at least ``tol`` for ``n_iter_no_change`` consecutive iterations,\n",
      " |      unless ``learning_rate`` is set to 'adaptive', convergence is\n",
      " |      considered to be reached and training stops.\n",
      " |  \n",
      " |  verbose : bool, default=False\n",
      " |      Whether to print progress messages to stdout.\n",
      " |  \n",
      " |  warm_start : bool, default=False\n",
      " |      When set to True, reuse the solution of the previous\n",
      " |      call to fit as initialization, otherwise, just erase the\n",
      " |      previous solution. See :term:`the Glossary <warm_start>`.\n",
      " |  \n",
      " |  momentum : float, default=0.9\n",
      " |      Momentum for gradient descent update. Should be between 0 and 1. Only\n",
      " |      used when solver='sgd'.\n",
      " |  \n",
      " |  nesterovs_momentum : bool, default=True\n",
      " |      Whether to use Nesterov's momentum. Only used when solver='sgd' and\n",
      " |      momentum > 0.\n",
      " |  \n",
      " |  early_stopping : bool, default=False\n",
      " |      Whether to use early stopping to terminate training when validation\n",
      " |      score is not improving. If set to true, it will automatically set\n",
      " |      aside 10% of training data as validation and terminate training when\n",
      " |      validation score is not improving by at least tol for\n",
      " |      ``n_iter_no_change`` consecutive epochs. The split is stratified,\n",
      " |      except in a multilabel setting.\n",
      " |      Only effective when solver='sgd' or 'adam'\n",
      " |  \n",
      " |  validation_fraction : float, default=0.1\n",
      " |      The proportion of training data to set aside as validation set for\n",
      " |      early stopping. Must be between 0 and 1.\n",
      " |      Only used if early_stopping is True\n",
      " |  \n",
      " |  beta_1 : float, default=0.9\n",
      " |      Exponential decay rate for estimates of first moment vector in adam,\n",
      " |      should be in [0, 1). Only used when solver='adam'\n",
      " |  \n",
      " |  beta_2 : float, default=0.999\n",
      " |      Exponential decay rate for estimates of second moment vector in adam,\n",
      " |      should be in [0, 1). Only used when solver='adam'\n",
      " |  \n",
      " |  epsilon : float, default=1e-8\n",
      " |      Value for numerical stability in adam. Only used when solver='adam'\n",
      " |  \n",
      " |  n_iter_no_change : int, default=10\n",
      " |      Maximum number of epochs to not meet ``tol`` improvement.\n",
      " |      Only effective when solver='sgd' or 'adam'\n",
      " |  \n",
      " |      .. versionadded:: 0.20\n",
      " |  \n",
      " |  max_fun : int, default=15000\n",
      " |      Only used when solver='lbfgs'. Maximum number of loss function calls.\n",
      " |      The solver iterates until convergence (determined by 'tol'), number\n",
      " |      of iterations reaches max_iter, or this number of loss function calls.\n",
      " |      Note that number of loss function calls will be greater than or equal\n",
      " |      to the number of iterations for the `MLPClassifier`.\n",
      " |  \n",
      " |      .. versionadded:: 0.22\n",
      " |  \n",
      " |  Attributes\n",
      " |  ----------\n",
      " |  classes_ : ndarray or list of ndarray of shape (n_classes,)\n",
      " |      Class labels for each output.\n",
      " |  \n",
      " |  loss_ : float\n",
      " |      The current loss computed with the loss function.\n",
      " |  \n",
      " |  best_loss_ : float\n",
      " |      The minimum loss reached by the solver throughout fitting.\n",
      " |  \n",
      " |  loss_curve_ : list of shape (`n_iter_`,)\n",
      " |      The ith element in the list represents the loss at the ith iteration.\n",
      " |  \n",
      " |  t_ : int\n",
      " |      The number of training samples seen by the solver during fitting.\n",
      " |  \n",
      " |  coefs_ : list of shape (n_layers - 1,)\n",
      " |      The ith element in the list represents the weight matrix corresponding\n",
      " |      to layer i.\n",
      " |  \n",
      " |  intercepts_ : list of shape (n_layers - 1,)\n",
      " |      The ith element in the list represents the bias vector corresponding to\n",
      " |      layer i + 1.\n",
      " |  \n",
      " |  n_iter_ : int\n",
      " |      The number of iterations the solver has ran.\n",
      " |  \n",
      " |  n_layers_ : int\n",
      " |      Number of layers.\n",
      " |  \n",
      " |  n_outputs_ : int\n",
      " |      Number of outputs.\n",
      " |  \n",
      " |  out_activation_ : str\n",
      " |      Name of the output activation function.\n",
      " |  \n",
      " |  Examples\n",
      " |  --------\n",
      " |  >>> from sklearn.neural_network import MLPClassifier\n",
      " |  >>> from sklearn.datasets import make_classification\n",
      " |  >>> from sklearn.model_selection import train_test_split\n",
      " |  >>> X, y = make_classification(n_samples=100, random_state=1)\n",
      " |  >>> X_train, X_test, y_train, y_test = train_test_split(X, y, stratify=y,\n",
      " |  ...                                                     random_state=1)\n",
      " |  >>> clf = MLPClassifier(random_state=1, max_iter=300).fit(X_train, y_train)\n",
      " |  >>> clf.predict_proba(X_test[:1])\n",
      " |  array([[0.038..., 0.961...]])\n",
      " |  >>> clf.predict(X_test[:5, :])\n",
      " |  array([1, 0, 1, 0, 1])\n",
      " |  >>> clf.score(X_test, y_test)\n",
      " |  0.8...\n",
      " |  \n",
      " |  Notes\n",
      " |  -----\n",
      " |  MLPClassifier trains iteratively since at each time step\n",
      " |  the partial derivatives of the loss function with respect to the model\n",
      " |  parameters are computed to update the parameters.\n",
      " |  \n",
      " |  It can also have a regularization term added to the loss function\n",
      " |  that shrinks model parameters to prevent overfitting.\n",
      " |  \n",
      " |  This implementation works with data represented as dense numpy arrays or\n",
      " |  sparse scipy arrays of floating point values.\n",
      " |  \n",
      " |  References\n",
      " |  ----------\n",
      " |  Hinton, Geoffrey E.\n",
      " |      \"Connectionist learning procedures.\" Artificial intelligence 40.1\n",
      " |      (1989): 185-234.\n",
      " |  \n",
      " |  Glorot, Xavier, and Yoshua Bengio. \"Understanding the difficulty of\n",
      " |      training deep feedforward neural networks.\" International Conference\n",
      " |      on Artificial Intelligence and Statistics. 2010.\n",
      " |  \n",
      " |  He, Kaiming, et al. \"Delving deep into rectifiers: Surpassing human-level\n",
      " |      performance on imagenet classification.\" arXiv preprint\n",
      " |      arXiv:1502.01852 (2015).\n",
      " |  \n",
      " |  Kingma, Diederik, and Jimmy Ba. \"Adam: A method for stochastic\n",
      " |      optimization.\" arXiv preprint arXiv:1412.6980 (2014).\n",
      " |  \n",
      " |  Method resolution order:\n",
      " |      MLPClassifier\n",
      " |      sklearn.base.ClassifierMixin\n",
      " |      BaseMultilayerPerceptron\n",
      " |      sklearn.base.BaseEstimator\n",
      " |      builtins.object\n",
      " |  \n",
      " |  Methods defined here:\n",
      " |  \n",
      " |  __init__(self, hidden_layer_sizes=(100,), activation='relu', *, solver='adam', alpha=0.0001, batch_size='auto', learning_rate='constant', learning_rate_init=0.001, power_t=0.5, max_iter=200, shuffle=True, random_state=None, tol=0.0001, verbose=False, warm_start=False, momentum=0.9, nesterovs_momentum=True, early_stopping=False, validation_fraction=0.1, beta_1=0.9, beta_2=0.999, epsilon=1e-08, n_iter_no_change=10, max_fun=15000)\n",
      " |      Initialize self.  See help(type(self)) for accurate signature.\n",
      " |  \n",
      " |  predict(self, X)\n",
      " |      Predict using the multi-layer perceptron classifier\n",
      " |      \n",
      " |      Parameters\n",
      " |      ----------\n",
      " |      X : {array-like, sparse matrix} of shape (n_samples, n_features)\n",
      " |          The input data.\n",
      " |      \n",
      " |      Returns\n",
      " |      -------\n",
      " |      y : ndarray, shape (n_samples,) or (n_samples, n_classes)\n",
      " |          The predicted classes.\n",
      " |  \n",
      " |  predict_log_proba(self, X)\n",
      " |      Return the log of probability estimates.\n",
      " |      \n",
      " |      Parameters\n",
      " |      ----------\n",
      " |      X : ndarray of shape (n_samples, n_features)\n",
      " |          The input data.\n",
      " |      \n",
      " |      Returns\n",
      " |      -------\n",
      " |      log_y_prob : ndarray of shape (n_samples, n_classes)\n",
      " |          The predicted log-probability of the sample for each class\n",
      " |          in the model, where classes are ordered as they are in\n",
      " |          `self.classes_`. Equivalent to log(predict_proba(X))\n",
      " |  \n",
      " |  predict_proba(self, X)\n",
      " |      Probability estimates.\n",
      " |      \n",
      " |      Parameters\n",
      " |      ----------\n",
      " |      X : {array-like, sparse matrix} of shape (n_samples, n_features)\n",
      " |          The input data.\n",
      " |      \n",
      " |      Returns\n",
      " |      -------\n",
      " |      y_prob : ndarray of shape (n_samples, n_classes)\n",
      " |          The predicted probability of the sample for each class in the\n",
      " |          model, where classes are ordered as they are in `self.classes_`.\n",
      " |  \n",
      " |  ----------------------------------------------------------------------\n",
      " |  Readonly properties defined here:\n",
      " |  \n",
      " |  partial_fit\n",
      " |      Update the model with a single iteration over the given data.\n",
      " |      \n",
      " |      Parameters\n",
      " |      ----------\n",
      " |      X : {array-like, sparse matrix} of shape (n_samples, n_features)\n",
      " |          The input data.\n",
      " |      \n",
      " |      y : array-like of shape (n_samples,)\n",
      " |          The target values.\n",
      " |      \n",
      " |      classes : array of shape (n_classes,), default=None\n",
      " |          Classes across all calls to partial_fit.\n",
      " |          Can be obtained via `np.unique(y_all)`, where y_all is the\n",
      " |          target vector of the entire dataset.\n",
      " |          This argument is required for the first call to partial_fit\n",
      " |          and can be omitted in the subsequent calls.\n",
      " |          Note that y doesn't need to contain all labels in `classes`.\n",
      " |      \n",
      " |      Returns\n",
      " |      -------\n",
      " |      self : returns a trained MLP model.\n",
      " |  \n",
      " |  ----------------------------------------------------------------------\n",
      " |  Data and other attributes defined here:\n",
      " |  \n",
      " |  __abstractmethods__ = frozenset()\n",
      " |  \n",
      " |  ----------------------------------------------------------------------\n",
      " |  Methods inherited from sklearn.base.ClassifierMixin:\n",
      " |  \n",
      " |  score(self, X, y, sample_weight=None)\n",
      " |      Return the mean accuracy on the given test data and labels.\n",
      " |      \n",
      " |      In multi-label classification, this is the subset accuracy\n",
      " |      which is a harsh metric since you require for each sample that\n",
      " |      each label set be correctly predicted.\n",
      " |      \n",
      " |      Parameters\n",
      " |      ----------\n",
      " |      X : array-like of shape (n_samples, n_features)\n",
      " |          Test samples.\n",
      " |      \n",
      " |      y : array-like of shape (n_samples,) or (n_samples, n_outputs)\n",
      " |          True labels for `X`.\n",
      " |      \n",
      " |      sample_weight : array-like of shape (n_samples,), default=None\n",
      " |          Sample weights.\n",
      " |      \n",
      " |      Returns\n",
      " |      -------\n",
      " |      score : float\n",
      " |          Mean accuracy of ``self.predict(X)`` wrt. `y`.\n",
      " |  \n",
      " |  ----------------------------------------------------------------------\n",
      " |  Data descriptors inherited from sklearn.base.ClassifierMixin:\n",
      " |  \n",
      " |  __dict__\n",
      " |      dictionary for instance variables (if defined)\n",
      " |  \n",
      " |  __weakref__\n",
      " |      list of weak references to the object (if defined)\n",
      " |  \n",
      " |  ----------------------------------------------------------------------\n",
      " |  Methods inherited from BaseMultilayerPerceptron:\n",
      " |  \n",
      " |  fit(self, X, y)\n",
      " |      Fit the model to data matrix X and target(s) y.\n",
      " |      \n",
      " |      Parameters\n",
      " |      ----------\n",
      " |      X : ndarray or sparse matrix of shape (n_samples, n_features)\n",
      " |          The input data.\n",
      " |      \n",
      " |      y : ndarray of shape (n_samples,) or (n_samples, n_outputs)\n",
      " |          The target values (class labels in classification, real numbers in\n",
      " |          regression).\n",
      " |      \n",
      " |      Returns\n",
      " |      -------\n",
      " |      self : returns a trained MLP model.\n",
      " |  \n",
      " |  ----------------------------------------------------------------------\n",
      " |  Methods inherited from sklearn.base.BaseEstimator:\n",
      " |  \n",
      " |  __getstate__(self)\n",
      " |  \n",
      " |  __repr__(self, N_CHAR_MAX=700)\n",
      " |      Return repr(self).\n",
      " |  \n",
      " |  __setstate__(self, state)\n",
      " |  \n",
      " |  get_params(self, deep=True)\n",
      " |      Get parameters for this estimator.\n",
      " |      \n",
      " |      Parameters\n",
      " |      ----------\n",
      " |      deep : bool, default=True\n",
      " |          If True, will return the parameters for this estimator and\n",
      " |          contained subobjects that are estimators.\n",
      " |      \n",
      " |      Returns\n",
      " |      -------\n",
      " |      params : dict\n",
      " |          Parameter names mapped to their values.\n",
      " |  \n",
      " |  set_params(self, **params)\n",
      " |      Set the parameters of this estimator.\n",
      " |      \n",
      " |      The method works on simple estimators as well as on nested objects\n",
      " |      (such as :class:`~sklearn.pipeline.Pipeline`). The latter have\n",
      " |      parameters of the form ``<component>__<parameter>`` so that it's\n",
      " |      possible to update each component of a nested object.\n",
      " |      \n",
      " |      Parameters\n",
      " |      ----------\n",
      " |      **params : dict\n",
      " |          Estimator parameters.\n",
      " |      \n",
      " |      Returns\n",
      " |      -------\n",
      " |      self : estimator instance\n",
      " |          Estimator instance.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "help(MLPClassifier)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-22T02:41:39.357949Z",
     "start_time": "2020-12-22T02:41:37.825688Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWsAAAD1CAYAAACWXdT/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8li6FKAAAgAElEQVR4nO3de3RU1dk/8O/O1ZAQEopC8KXyNlpRCWiRVKUFRC5FVED0FQoExJ9VaVmKUYI/Ly31Z1uwpGhAkbYULNS8VgVF1IQmXIRSg7GQiFBbV7GoR5QkkARCyGX//picMEnOTM7MnPt8P2tlGZLJzAYzz3nOs/d+tpBSgoiInC3G7gEQEVH3GKyJiFyAwZqIyAUYrImIXIDBmojIBeLMeNLU1N6yb98BZjw1EZFn/etfFcellOdrfc+UYN237wDk579txlMTEXnWLbdc+Gmg77EMQkTkAgzWREQuwGBNROQCDNZERC7AYE1E5AIM1kRELsBgTUTkAgzWREQuwGBNROQCDNZERC7AYE1E5AB9K3cH/T6DNRGRCzBYExHZrLusGmCwJiKylRqos7N7B30cgzURkc26C9QAgzURkW36Vu7WFagBBmsiIlvoqVP7Y7AmIrKY3jq1PwZrIiIbhBKoAQZrIiJLhVr+UDFYExFZJJzyh4rBmojIQuEEagCIM3gcRB0oyhG8vmUtduzchIb6E0hKScPoUVMx+eZ5yMgYaPPoiKwTyjI9LcysyTTl5aW4P3cS9lbVIW3GUgx4aBPSZizF3qo63J87CeXlpXYPkcgS4dap/TFYkykU5QiWLl+AtCmPInVkDuLTMyBiYhGfnoHUkTlIm/Ioli5fAEU5Yus4icwWSZ3aH4M1meL1LWuRlDUeiRdepvn9xAsvQ1LWOLzx5h8sHhmR9SIN1ACDNZlkx85NSMoaF/QxSVnjsWPnJotGRGQ9I8ofKgZrMkVD/QnE9bog6GPiUs/H6foTFo2IyFpGlT9UXA3icorizNUWSSlpaD75FeLTMwI+prn2a/RISbNwVETWMipQAwzWrlZeXoqlyxcgKWs80mYsRZ9eF6D55FfYW7kNpbmTkJdbgGHDxpg6BkXRvlgMv/oGVFRuQ/zInIA/21BZjNGjppo6PiI7RLpMTwuDtUspyrnVFv6TePHpGYgfmYPEzOFYunwBnlm+1bQMO9jF4nRFMVpbmpGYOVxzkrHx80NoqNyGW5ZvNWVsRHYxsk7tjzVrl7J7tYWiBF+alz71MYiYGFS/9nPU7lqPphoFsqUZTTUKanetx4nNTyEvt4AbY8hTjK5T+2Nm7VI7dvo2mASTlDUeOwoX4567lxj++nouFslDf4ChPWKQ3KMndhQuxun6E+jRVia5xcSMn8hOZgRqgMHatRrqT6CPjast9F4s9hUuxksbKky5YBA5iVnlDxXLIC6lrrYIxszVFlyaR3SOmeUPFYO1S40eNRUNlduCPsbM1RZ2XyyInMKKQA0wWLvW5JvnoaGyGI2fH9L8fvtqi5vuNOX17b5YEDmJ2YEaYLB2rYyMgcjLLcCJzU/ZstrC7osFkROYXaf2xwlGFxs2bAyeWb4Vb7z5B8tXW6gXC98663FIyhqPuNTz0Vz7NRoqi9FQuY1L88jTrCp/qISU0vAnveSSoTI//23Dn5ecR1GO+C4WOzd1vFjcdCcDNXmaGbsUxbhx5VLKqzW/x2BNRBQaMwI1EDxYs2ZNRBQCK+vU/liz9hBFcWYHPiKvsLpO7Y+ZtUfwvEMia9gRqAEGa09QFJ53SGQ2u8ofKgZrD7C7Ax+R19lZ/lAxWHsAzzskMp+dgRpgsPYENlUiMo9Zy/RCxWDtAWyqRGQOu+vU/hisPYBNlYiM54Q6tT8Gaw9gUyUiczglUAMM1p5gdwc+Iq9xUvlDxR2MHmFnBz4iL3Fa+UPFYO0hGRkDcc/dS3jeIVGEnBaoAZZBiIjaOWWZnhYGayIiOLNO7Y9lEHIcRWH3QLKWU+vU/phZk6OweyDZxcmBGmBmTQ6iKOe6B/o3pYpPz0D8yBwkZg7H0uUL8IxBq1sUhRk8ObtO7Y+ZNTmGld0DmcET4Pw6tT8GaxspyhGsXvMEps/MwuTJAzB9ZhZWr3kiavtOW9U9UFHY/5vcUaf2x2BtE6MzO0Vxf+C3qnsg+3+Tyi2BGmCwtoWiGJvZmXVLH7tnC36VOxY1NcE7+hnFqu6B7P9Nbip/qBisbWBkZqco5t3Sb3j3NRw6+g8UvZCnaxyRZvZWdQ9k/+/o5rbyh4rB2gZGZnZm3dLH7tmCd/bvRElOEt76+w7E/fXNgI81KrO3qnsg+39HL7cGaoDB2hZGZnZm3dJvePc1zBkai6syYjH3yjhs2PWq5q2johiX2VvVPZD9v6ObGwM1wGBtCyMzOzNu6aurj+Gtv+/Aomt9vx6Lro3BOwd2oqruRJeAHWlmrygdyydP59+P7w4fh6E9YnCycDGO5k/DycLFuK5PLzyzfCuGDRuj++8RCPt/Ryc31qn9MVjbwMjMzoxb+qI1izFnSBwyevp+PTJ6xiBnSCy2/WNrl8dGktkHKp9UnBH423tFeGjhCry++T94aUMF7rl7iWEbVdj/O/q4ufyh4g5GG0y+eR5KcychMXO4Zkbantkt7xocOxs9air2Vm5D/MicgI8JFPgVpesOvmu+OwHvf7Adh+cndnjsomtjMPiFIoy7dBL6Vu7Gsazv+Z67/gT6hJHZK4q1uxU7Y//v6OGFQA0ws7aFkZlduLf0gbLa9z8qw8zBoj2rbh9zp+xafQOEm9k7Ya2z2v/7pQ0VpmTw5BxuD9QAg7Vt1Mzuuj69IqrNhhP4FUV7UlDEJ6K5+igeH5mg+VqLro3B+uIifPNi35/7Vu4Ou6Tj1bXOiuL+zUle4vY6tT+WQWxk1Mkuod7SB8pqz7xXiLlD47pk1e3jbcuulxVuRP78BSgrq8Zd3x6C0jU/DbmkE275xMnKy0uxdPkCJGWNR9qMpejT6wI0n/wKeyu3oTR3EvJyCwyZICV9vFL+UDFYe0QogX/HTl/Jo7PWLw/juc9O47n3gv/8iEEHAbS9CcqAX956Lx557SkkZY1DUtZ4xKWej+bar9FQWYyGym2aJR21fBKfnhHwdaxc66wokXXgUxR7a/DUkdcCNcBgHZUCZbWpM59FatvnsqUZn+XfipaioqDP5XszXIWNP1qCtf+s1D1ZF8nEqNGMyIhDqcHzjExreClQAwzWUUlvVtszpZeu51Mz7Ce+0U93INJaEdNUo6Dugzdx6qOdaG04CRGbgPqRk6EoR0zLRhXFmIw40N2Kv6Ss8dhRuJjB2mReqlP74wRjFNIzKdhYWYRZN1yv+znVLEbvG6XzxGjdgWJ8+cdciLh49Jv1NL750GZkzFuJigaY2l/aqFUp7DfiDF4sf6gYrKOQnuV+jZXFWDhlckjPG2rAVidGhyYJ1JS8gAumPY70UXMt7S9t1KoUozYnKQpXk0TKi4EaYLCOSsGW+9XvWof6zU/i5UfykNm/f8jP7R+w9QTtjIyB6NGjJ9KunmzLmuvTdTWoK38DRwtm4dNlt+BowSxUl/wWTTVK+2P0ZMRG7Erl6TWRccvxXOFisI5SXdZ5L78VpwoXYXq/VuxfVYCJ2dlhP3d2du+Qsmy71lyXl5dCxCdAxCe2lV42od+spyHiEvDlH3PR8Mn7APRlxOFuTlIUXyZ9xw8HY8mSHJxtaUFzUyMA8PSaEHi1Tu1PSCkNf9JLLhkq8/PfNvx5vUpRfMvGtu94DQ2nTkDEJkDKVpx3Xg+Muf42Sw5wNSsrKSurbv9c3aLe2eTJAzDgoU0QMbEBn0e2NONo/jS8vvk/hoxLUY7g/txJXSYWVY2fH8JXrz6JfrOXo6GyGNf16dXtxOC5VSWBlzD6ryrxX4WSlDUOcW2rUOorilF/oAh9Jj2IpMyr2x9fu2u9rnFEGy/VqcW4ceVSyqu1vsfM2mbqre9fj9ci/YfL2ifWUodPwdmWVuz618euvgXWk2Xb0V9az8RiytDxOPHuH3V34AtlV6qiBG4tmz5qDi6Y9jiOb83vUI5x445Oq3ghUHeHS/dspCiBl42lj5qDHhdn46tXn0Tv8feZuqHCiltI9c1UVuZ7Lf8s244113qW2qUMmQBl7U/wxKO/0/3vrndzkt6LRd0Hb6L3DXcD4GoSLV6vU/tjZm0CRdE3o6/3Ddv4+WHTmxpZ9QuvlWXb0V9a71I7tDaZskVcT50+ZcgEnDq0s/3PPL2mo2ioU/tjsDZYKDP6obxhvXQLrJZG1BUjdvSX1l96STfsNf3pvVi0nq499zM8vaadl+rUejFYG0hRQjviKpQ3rBdvgf2z7BsTEgzpQqiX3Ud76b1YxPTwNQDg6TVdRVOgBsKsWQshUqSU9UYPxu1C7Q+hd9t3TI9Uz94Cn6tlV+PK45/hymvGWbLawcgDIMKhp05ff6AISd8ajtpd6wM2xIpG0VSn9hduZv2RoaPwiFDXC+vJ7uoripB82SjP3wJ3XjVidj3S7qO99NTp68q3oPXf+0y7u3CjaKtT+wuYWQshHgz0LQAp5gzH3ULt0awnu6s/UIze4+9DfckLpmV5TuKfaatvzEDrsyNl59Fe6sUi2LrsJx79HQO0n2isU/sLuClGCHEGwNMAmjW+vVBKGfCePFo3xUyfmYW0GUuDljWaahScLFyMlzZUADi3MeK8wWPRY8iE9jds/YEi1B8oQo+BQ9H82YemNa53+htA3VRjVsC2m6Ic8V0sdm7qeLG46U6WPDqJhvJHsE0xwYL1XwEskFKWa3zvqJRyQKAXjNZgvXrNE9hbVYfUIHVIrV1oiuJ7w5ZufxUNp05CxMVDtrYiKSkZY66/zdQ3rlveAF4P2hRcpL+nSlUV7lz2C6zLexT9ejv39z1YsA42wXgngKoA39N8smgX7qSVUcd7eVl2dm9LSiPkPEbUqZcVbkTZoQ/bj6Rzo4ATjFLKf0gpjwf43jHzhuRedk9aeZ3Vk5BkPyPKdEpVFdYXF6NkdhLWFxfhy+rq7n/IgaJinbWiWNcj2KhTyykwBm3vq64+hp8/Pg1VdSciLtMtK9yIOUNjcVVGbPuBz27k+a57gTqbNVRuQ0NlcdSfOO2WmnUwejr7WU1RIjuAN9qtXfMIdv5lA+ZOvCmisoVSVYUr5s3BwXvjkdEzBkpdKwa/0ISDa190ZO06arvuKUpoOwrJnULtn202HiIQmerqY9heUoiSnMjLFmpWndHTF+oyesa4NrvuNlgLIb4thCgRQnzY9uchQojHzB9a5Iw6X4/coXPPETsoChOESBWtWYw5Q+IiLluotepF13YMc4uujXFl7VpPZv1bAI8AaAIAKWUFgOlmDsoodp1AQvayM8vWkyAkXHY97pk/iucraqiuPoa3/r4Di0f4QlMkgbVzVq1ya3atJ1j3kFKWdfqa1kYZx3HbidOKYu1hqU4oGZjFroCtq5PiVTciJjGFpRENRWsW484r4yIuWwTKqlVuzK71BOvjQohMABIAhBC3AVCC/4gzWH0CiaKEH2ztqnOaObmoVFXhB3m5tr0h7AjYujspNtSyNNJJ7J4teOvvOwwpWwTKqlVuzK71BOsfA3gBwCAhxOcAHgBwr6mjMoiVbTAjCbaK4s06p/9GBLtYHbBDbX0KcO4E8P3/2fDuax2yalU4gXXf4Y+wYu9piCW1AT9W7D2NskMHjf6rmCZosBZCxAC4Wko5FsD5AAZJKb8npfzUktFFyKoTSBQlsmDrxYlQJ21EsDJgh9JJ0V+0z50cr63BO/t3Gla22F2wGnLbtm4/dhesDvgcdt8ZdhY0WEspWwEsavv8lJSyzpJR6aAo3ZccrNpRGGmw9eJEqNM2Ili1llxPglB/oBg9v3NTh687ae7Ean0rd2PbP7Y6rmzhhDtDf3rKIH8RQjwkhBgghOitfpg+siBCKTlYsaMw0mDrtonQ7nSe3HHKZI66rM9MwRKEmp3r8NWrT6LPpAe7dGb06uES3VH/fzitbOGkO0OVnpNi7mj774/9viYBfMv44XRPUQKfCB4/MgeJmcO7nARudqOkUPtYd6b3xBi3vJmDbURwQhOdvpW7Td3p2LlP9qm6GsQkJCE5ayz6zV6u+f/Z64dLaPHv+7E7O3A5wg4d7wylI353u82spZT/rfFhS6AGnFnfjXTVid3nARrJ6RsRrKpfqwnCSxsq8MLqd5EQH4/kQd/TDNTRfL6iE1sdOPXOUM8OxhytDysGp8Wo+q6iGLemOdJga9VEqBWM3ohgxiSP1QGC3Ri7cnJPGqduUddTsx7u9/F9AD8DcIuJYwrKiPqu0WuaIw22dryZzXizmLERwaxJHivq1/7YjfEcJ2/GcvKdYchd94QQaQAKpZQ/CPQYM7vuhXN0lj9FOYL7cyd1qXmrGj8/hBObn+pQ89bjXHc/7fP09HT3UxTrjngyI1gvXPUscKwEvxkfeCpkYXEzRL+xuup/ase0kllxGLux2fBOaf6nzygKu+RZwenHyAX7HQ7ldzdcYR3rFfDJhIgH8KGU8tJAjzEzWId7dJZRPx+Mojj3PD1F6RiMknv0RM64MXhw6hRk9u9vyGt8b8G92HP4k24fN2JQZtD1rSr/N45Zb5Sysmq8dfYs2+haxMnlj87tVLt834L2qhEFayHEFrRtNYevbHI5gD9LKfMC/YyZwVpRIsuMI83MraIoxmV6gXp6N35YjMaKIrz8SB4mZmeb8dcImxF9iPWcu/fJF18g674F6HXr44beaVFXTg7UgPF3huEI9wxG1a/9Pm8G8KmU8jNDRhYGtb7bXckh0Bsr0mV2VvAPrmkzlqJPW3DdW7kNpbmTQsr0FCXIUsfvz0H8t7LxP798EvtXFRiWYRvBiOV/es7dy9+0GUlDJuhaXcQzMoNTlMAJxpXHbQsZuu07/BH2HD6NFXuDP27EIHu2qOvJrJd2zqK1vubPipNiFCW8koPTM2tFMbamrqfsU79rHab3a8XK+feFP3ADBbodDSW71lvvTpt2O5KnP+3Y3we30Lp7a1Q+xqm3lqG5oQG/vP0neGD2DXYP0/EiPSlGa53cxMiGFDn/dayvb/4PXtpQgXvuXtJtAHP6mmaj15HrWeqYmDUBG0q2hzxWsxix/E/vdvfaulpP7R61g6Jo98ZpPrwdsaercN7AIXj0lZUY9cBPbF+r7GYBg7UQ4j4hRCWAS4UQFX4f/wbg2hTD6Wuaje4TonepY139Sd1jNJMRy/9C2dSQ2jO12w1NZz47iLiEJMv6jLuNVoLRXF+Nhg//gu05STh7pByxqd9AxScf275W2c2CZdZ/AnAzgDfa/qt+DJNSzrJgbKZw+gYFo/uE6N1d2TOll+4xmsmIPsShbGqYOeZ6NH5YHPC5Gj55H19vego9hk7geYoBaCUYZ94rxNyhvqO5bh8Ug+bqoyjN6WH7WmU3CzjBKKU8CeAkgBkAIIS4AMB5AFKEEClSyv9YM0Tjde7d0KHmbfOsv9F9QkaPmoq9ldsQH6Rm3VhZhFk3XB/yWM0Q6SSPmlUfvDe+w9cXXRuDwS8UYdH0mR1q1w9OnYIXf7wA8d/K7lJ6aqpRcPzNX6Pv7Ut096GJRp0n7dWs+rH5CQCABNGMuUOc1WfDjbpdDSKEuBlAPoD+AL4CcBGAQwCuMHdo5jK7uVO49ATXUGrqk2+eh9LcSUjMHB5wwrKxshgL7y0Ie8xG0rP+Ohg99W7/QJHZvz9efiQP//PLJ9GUNR6JWRPaVxdVvflrJA8Zx5Ui3eicYKhZtbrk8pWPmnBwfgqAwBdN6p6eCcb/B+AaAB9LKf8bwA0A/mbqqKKY0TX1YGWf+l3rUL/5Sbz8SJ6jlu2FK9x698TsbOxfVYDp/VpxqnARji6/FScLF6O1+jP0vPLGoK/ptj7jZvCftG/PqtsOvF225yzmDE1wXJ8NN9ITrJuklFUAYoQQMVLK7QA0l5ZQ5MyoqWv1pajd+DCm92vF/lUFjtsQE65I6t2Z/ftj5fz7cOLVl9FaXITSvOfQ3NjAlSI63PXtIThz4B00fn6oS1a9/sBZLBqR0OHxTuiz4UZ6NsWcEEKkAHgXwEYhxFcATpk7rOhmRk29c9nH6bvJwmH0pobkHj091WfcSP7NmLInXo5Xv5GH255aAtncgMd+kgSga1atclp/c7fQsykmGUADfFn4TAC9AGxsy7Y1WbEphiLjxWBttNt/+huUyBRT+si4VYcg3en3Z96yX6LHyT1YOTERSl0rrniuHgfnp9jWZ8ONItoUI6U8BWAAgNFSyvUAfgfgrLFDJHKeGddMaL+912L3mnwr9a3c3aFjntaF/uOjn2JVWSPEklpc/Gw9pg/WbogEsHYdDj2rQe4G8CMAvQFkArgQwGr4JhptpyhsbUna9DRyCubWiZfjF8fn45HXngqrD40XBMukO1NX8ixc9SzWvr0Fz7/fhOffbwr6M3b12XAjPTXrHwPIBvAeAEgp/9m25tp2RjY8Iu/R08ipO9ddehU2/mgJ1v6z0nFr8s0SSoDuTF2Rs2NOsik9yKOZnmDdKKU8K4QAAAgh4nCuZaptFCX0g3MpevifTj12Y/jrerOzewNlwD2jb/N8XTqSIK1y4kGzXqFn6d5OIcT/BZAkhBgH4M8Atpg7rO458eBccg69jZz0cvJRVJHSU4/Ww6kHzXqFnmC9GMDXACoB3APgLQCPmTkoPYxueETeYXTQ8OqqGaOCtMqpB816RbCue98EACllq5Tyt1LK26WUt7V9bnsZxOiGR+Qdywo34rbLYnDn6w34sr7VsKDhleza6CANOPugWa8IlllvVj8RQrxqwVhCorebXDRuWIhmatCAbEbZ5y1Ytse3ytSO7FpRjmD1micsb61aXX0MP398Gmpqzr0/1ACtrq83KkirjOhBTsEFC9bC7/NvmT2QUDn9EAGyh5pVv/JRE0pykrH+QJNh2XV2dm/d2XV5eSnuz52EvVV1lrdW3fzKCnzycRk2vbLClCy6MyN6kFP3ggVrGeBzR3D6IQJkPf+ses7QhLbJxXjDsmvd41C0T04BgOamRpxtacGSJTm444eDDc+0q6uPYXtJIUpmJ2HHX17CNy82L0irjOhBTt0LtnRvqBCiFr4MO6ntc7T9WUopU00fXRCRHpxL3tOeVR9sPNeSc0QCBj9/CotGJBjSkyI7uzfKynbjWNb3Aj5Ga6VSwyfv4/jWfKQMnYB+s/PbT5g3ck9A38rdeGnr7zFniK/p/9wrrVk65/SDZr0i2OEDsVYOJBxOPkSArLfv8EfY988G3HVVfIcVCXdcEYeLVtTjbIvvcUYEjb6VgQP2jp2+koeqqUbB8a35uGDa44bvCfAvyxyvrcE7+3e2H7wQae9ovTtAI+1BTvp028gpHGzk5HxebORkxKnoepWVVXcJ1oria33w9jt/hGxpRkxSKpIvH4XWM6cQm5KO9FFzAj5fqA2htDawLFz1LHCsBL8Zfy4HW1jcDNFvbFjZ9cJVz2J90VbM/cFN3NhikUhPNyeynVJVhR/k5QatN1u9IsE/YPpPKGbMW4VvPrQJ/WY9DRGXgNOHdyFlyPigz6V3T0CgCUOjl8757wDl5KAzMFiTK/j3+dBi9YoE/7sSRdGeUIxPz0D6qDmQLU0R7QnQs+zO6AuV0TtAKXIM1uR4erI8PSsSZmeZk1131/ogJqlXWHsC9C67M/pCxW3jzsRgTY6nJ8vbd/gjrNh7GmJJbcCPZ/52GmWHjFuRoAbPXdtfCdr6IPnyUaivKAr6XP57AjoH6QGZMmgJyOilc9w27kx6uu4R2UbN8rpb4RBoRYI66VgyKw5jNzbjlSW/MHR82dm9cep0HXoHKXP0/M5N+PKPuehx8XcDnjB/5kAR5t37pObEb3etXo1cOtf531vFU8ntx2Adhdy0EiRYlqdnhYIVLTu7O6sxPj0DaaPm4tj/Poa0q2/puCegoghnKorxi9vm49aJl3f5WT2tXo1cOqen9s2VIfZgGYRComdVhpGvFckKB6tqrznjxuBMRXHQx8iTCkZ9fzJuEKdRu/FhHF1+K2o3PowfZrSi8vkCPDBb++AlKyf6uG3c2RisKSTdrcow+rUiWeFgVe31walTcPbDoqCtD84cKMKCK0fg4Uk5qNv8Z7QWF6Fu85+xcv59yOzfX/PnrJ7o01P7nnG5xM/W/d6U16fgGKypW2o2feCTTyxbextplmdly87M/v3x8iN5qN/8JOp3rUNTjQLZ0oymGgX1u9ahfvOTePXRPNw68fKQyk9WT/TpmaRdte8siva9Z8rrU3AM1tQtNZu++9e/ar8lnzYoBt+/f76tWV6wwGX1BpmJ2dnYv6oA0/u14lThInyWfytOFS7C9H6t2L+qABOzs0N6Pjv6Q+8uWA25bZvmxxeFhUjvkYgPfpSM2oYzLIXYgBOMFJQaNAqnJWJK4RG8fouvQRJkM76uqcLP1v0eqx982PDXjWSFQ6AVDSqzVjZk9u+PlfPvw8r590X8XE6b6Os8UfvEut/jP8e+DPvkeAode4NEoVBWg6j9JtDajOZWiYIbk6DUteKK5+pRkpOMUevP4OMX/+SoN6xWj4wuj4mgZ4bZAvU4af++Cb1OQhmPUteKQasaEBsj2TfEYMF6gzCzpoDUDLV0VizGvHi2ve3osj1n2/tFz8qKMS27DpfbW3aGUgKyIlBqjadVtmDH7OSITo6n0DBYU0Dqm3T9AV8zfzWrWn/gXOB+fGQiLl1Vgp/Nvcsxb9jdBatd3THOSRcbrZLSsj1ncddV8bgqIxazs6zpmU0M1hRAd1m1/woFM7Jrvb2Ug4092EYSJ3NSf+jOWXXni3XeddzZaBWuBiFNwbLqRSMSOjz28ZGJ+FNpiaErBCJZz82OccbQWpGidbEOtUGWlRurvITBOsromVz0f5Pu+6IFK947C7GkFgOfqcf0wV0nvTJ6xmC2gUExkl7K7BhnnEBZdeeLdd512v/GgYKylRurvITBmrrwf5PunpcM+dNUfPFgCpLjfVm0lse+n2BYUIwkM2bHOON03iQT7GKt9W+sFZR5qEH4GKypC62dbIHeqCqjgmIkmbHejSS8DdfHf5PMF4WFSCwqdPUAAAkTSURBVE5MDHix1vo31grKLFGFj8GautDayTb8kkw8/35T0K3IK/ZG3i86ksxY765F3oaHLtQdpVpBmSWqyHA1COlixQqFSHop6921OGfCja5eKWKXUJYTBupBXt/QEFG722jHYE2OEckWa72Zn39/k879rSNZLuh1oVysF656tktQnp0Vi7XbS/GPHyd1eCwPNdCPZRByhEi77OnpGLdi72kc/PTTgLfhLI9ELtD/x7zrYgDZAiE6Pp4TwPoxWEcR9Vw/J4q0y16wjnHqxwNTbsaPrk7SvA3nKgVjBLs7mndVPJbtOdvlZ1i71odlkCjj1OO8zN5i3V09XK2nmnn8l9d1N2+QNyIRg58/hUUjEtAv5VwwD7V2Ha3lKgZrcgSzJzCDZXyd66mso4ZHz93RHVfE4aIV9Tjb0vX7ei/E3R0g7FUM1uR53WZ818Vg7Qfn6qlcpRAe/XdHmWFfnN3e9yUSDNbkeXoyPrWemj/hPADMrsNhxfJOK06rdyoGa/I83RnfgNj2z5ldO0+g9dvRckFlsI4SoZwO4zWBMr7uTmSJtmDgdMF2t0bDBZVL9yhqRbpckKxjxwHCTsPMmqKWk05koeCcdoCwHRisKWo56UQWCsyu0+qdhmUQMhxbkJKRWK7yYWZNhovWTQudRetOO6OxXOXDYE2GiuZNC51ZedHy8oWB5SoflkHIUDwJxMfqxlDsGOh9ngnWinIEq9c8gekzszB58gBMn5mF1WuegKIcsXVc0YQngZxj5UWLHQOjgyeCdXl5Ke7PnYS9VXVIm7EUAx7ahLQZS7G3qg73505CeXmp3UOMCtF4WK3WZKrVFy3ezUQH1wdrRTmCpcsXIG3Ko0gdmYP49AyImFjEp2cgdWQO0qY8iqXLFzDDNlk0bFrQCsxa5QcrL1p6LwxcoeN+rg/Wr29Zi6Ss8Ui88DLN7ydeeBmSssbhjTf/YPHIoovew2rdrHNg1io/WH3R0nthYE3b/VwfrHfs3ISkrHFBH5OUNR47dm6yaETRJ9IjudxAKzBrlR+svGjpvTCwpu0Nrg/WDfUnENfrgqCPiUs9H6frT1g0ougTDZsWOgfmn677fZfywx/eeQfri4ssu2jpvTCwpu0Nrl9nnZSShuaTXyE+PSPgY5prv0aPlDQLRxVdvL5pQas156WrSjB9cEKH8sPF6a245r/0XbQiXXetdwv2nAk3RnVbUS9xfbAePWoq9lZuQ/zInICPaagsxuhRUy0cVXTx+qYFrbrwrKwYtLQ2AzgXLGNEK1aWNWNlWWPQ5zPioqX3bubuX/8qYE374Tt+6NmNNF7k+mA9+eZ5KM2dhMTM4ZqTjI2fH0JD5TbcsnyrDaMjtwuUwT4+0nf465LrW9sPf33v/6RgYXEzRL+xpu9Y1Hs3k5r4KV6/JbnD19Ts+tSZM2wL4CKur1lnZAxEXm4BTmx+CrW71qOpRoFsaUZTjYLaXetxYvNTyMstQEbGQLuHSi4UvC7sOwrMn1Zd2oxlc7sLVkNu2xb044EpN2Ped5I0xz5tUAz+VFrCSUcXcX2wBoBhw8bgmeVbcV2fXjhZuBhH86fhZOFiXNenF55ZvhXDho2xe4jkQt2uchmRgPUHmvBlfWv717QmU+1YNtfd2CGbMTsrhpOOLuL6MogqI2Mg7rl7Ce65e4ndQyGXCdQESU9d+I4r4nDRinqcben4PbUubVdjq2BjV+pa8cpHTTg4PwUAJx3dwjPBmihcgbrj6V/lkhlwktWu07iDjT0hFrjrqnjNSUfWrp1LSCkNf9JLLhkq8/PfNvx5KXzRfGBuMOqhuSWz4jB2YzMOrn3RsOyy84G8Sl0rBr/QZOhrRDqm9q87YGwEiHHjyqWUV2t9zxM1a6JwmblhxImNraKhLYBXMVhT1DKzO54TG1tFQ1sAL2OwpqhlZubrxAw2GtoCeBknGCkqBdrsYsTKCKeexu31tgBex2BNUUlP5hvuyohQMlgrV194vS2A1zFYU9QxO/NlBktmYLCmqGN25ssMlszAYE1Rh5kvuRGDNUUdZr7kRly6R0TkAgzWREQuwGBNROQCDNZERC7AYE1E5AIM1kRELsBgTUTkAgzWREQuwGBNROQCDNZERC7AYB0leP4ikbsxWBMRuQCDNRGRCzBYExG5AIN1lCgr44nVRG7GYE1E5AIM1kRELsBgTUTkAgzWREQuwGBNROQCDNZERC7AYE1E5AJCSmn8kwrxNYBPDX9iIiJvu0hKeb7WN0wJ1kREZCyWQYiIXIDBmojIBRisiYhcgMGaHEkI0SKE2O/3MTCM50gTQsw3fnTtzy+EEM8KIf4lhKgQQnzHrNciirN7AEQBNEgpr4zwOdIAzAfwXCg/JISIlVK26HjoRACXtH18F8Dzbf8lMhwza3INIUSsEOJpIcS+tkz2nravpwghSoQQHwghKoUQk9t+5FcAMtsy86eFEKOFEG/6Pd9KIcTcts+PCCGWCiE+AHC7ECJTCPGOEKJcCPGuEGKQxpAmA3hR+vwNQJoQIsPUfwSKWsysyamShBD72z7/t5RyKoC7AJyUUg4XQiQC2COEKAZwFMBUKWWtEKIPgL8JId4AsBjAYDVDF0KM7uY1q6SU32l7bAmAe6WU/xRCfBe+7HxMp8df2Pbaqs/avqaE+XcmCojBmpxKqwwyHsAQIcRtbX/uBV8J4jMAvxBCjATQCl/A7BvGa/4v4MvUAVwH4M9CCPV7iWE8H5FhGKzJTQSABVLKog5f9JUyzgcwTErZJIQ4AuA8jZ9vRsfSX+fHnGr7bwyAEzpq5p8DGOD35/9q+xqR4VizJjcpAnCfECIeAIQQ3xZCJMOXYX/VFqivB3BR2+PrAPT0+/lPAVwuhEgUQqQBuEHrRaSUtQD+LYS4ve11hBBiqMZD3wCQ0/b9a+Ar0bAEQqZgZk1u8jsAAwF8IHz1ia8BTAGwEcAWIUQlgPcBHAYAKWWVEGKPEOJDAG9LKR8WQrwM4EMA/wbw9yCvNRPA80KIxwDEAygEcKDTY94CcCOAfwE4DeBOQ/6WRBrYG4SIyAVYBiEicgEGayIiF2CwJiJyAQZrIiIXYLAmInIBBmsiIhdgsCYicoH/D3fw3XGXXMqoAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.neural_network import MLPClassifier\n",
    "from sklearn.datasets import make_moons\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X, y = make_moons(n_samples=100, noise=0.25, random_state=3)\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, stratify=y, random_state=42)\n",
    "\n",
    "mlp = MLPClassifier(solver='lbfgs', random_state=0).fit(X_train, y_train)\n",
    "\n",
    "mglearn.plots.plot_2d_separator(mlp, X_train, fill=True, alpha=.3)\n",
    "mglearn.discrete_scatter(X_train[:, 0], X_train[:, 1], y_train)\n",
    "plt.xlabel(\"Feature 0\")\n",
    "plt.ylabel(\"Feature 1\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We are using 100 hidden nodes which is quite a log for a small dataset. We can change it by setting hidden_layer_sizes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-22T02:41:47.167987Z",
     "start_time": "2020-12-22T02:41:46.727604Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.neural_network import MLPClassifier\n",
    "from sklearn.datasets import make_moons\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X, y = make_moons(n_samples=100, noise=0.25, random_state=3)\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, stratify=y, random_state=42)\n",
    "\n",
    "mlp = MLPClassifier(solver='lbfgs', hidden_layer_sizes = [10], max_iter=1000, random_state=0).fit(X_train, y_train)\n",
    "\n",
    "mglearn.plots.plot_2d_separator(mlp, X_train, fill=True, alpha=.3)\n",
    "mglearn.discrete_scatter(X_train[:, 0], X_train[:, 1], y_train)\n",
    "plt.xlabel(\"Feature 0\")\n",
    "plt.ylabel(\"Feature 1\");"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-22T02:43:32.944908Z",
     "start_time": "2020-12-22T02:43:32.491723Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.neural_network import MLPClassifier\n",
    "from sklearn.datasets import make_moons\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X, y = make_moons(n_samples=100, noise=0.25, random_state=3)\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, stratify=y, random_state=42)\n",
    "\n",
    "mlp = MLPClassifier(solver='lbfgs', hidden_layer_sizes = [10, 10], max_iter=1000, random_state=0).fit(X_train, y_train)\n",
    "\n",
    "mglearn.plots.plot_2d_separator(mlp, X_train, fill=True, alpha=.3)\n",
    "mglearn.discrete_scatter(X_train[:, 0], X_train[:, 1], y_train)\n",
    "plt.xlabel(\"Feature 0\")\n",
    "plt.ylabel(\"Feature 1\");"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-22T02:44:36.491283Z",
     "start_time": "2020-12-22T02:44:35.649175Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.neural_network import MLPClassifier\n",
    "from sklearn.datasets import make_moons\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X, y = make_moons(n_samples=100, noise=0.25, random_state=3)\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, stratify=y, random_state=42)\n",
    "\n",
    "mlp = MLPClassifier(solver='lbfgs', activation = 'tanh', hidden_layer_sizes = [10, 10], max_iter=1000, random_state=0).fit(X_train, y_train)\n",
    "\n",
    "mglearn.plots.plot_2d_separator(mlp, X_train, fill=True, alpha=.3)\n",
    "mglearn.discrete_scatter(X_train[:, 0], X_train[:, 1], y_train)\n",
    "plt.xlabel(\"Feature 0\")\n",
    "plt.ylabel(\"Feature 1\");"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-22T02:47:12.407727Z",
     "start_time": "2020-12-22T02:46:55.812008Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x576 with 8 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(2, 4, figsize=(20, 8))\n",
    "for axx, n_hidden_nodes in zip(axes, [10, 100]):\n",
    "    for ax, alpha in zip(axx, [0.0001, 0.01, 0.1, 1]):\n",
    "        mlp = MLPClassifier(solver='lbfgs', random_state=0, max_iter=2000, hidden_layer_sizes=[n_hidden_nodes, n_hidden_nodes], alpha=alpha)\n",
    "        mlp.fit(X_train, y_train)\n",
    "        \n",
    "        mglearn.plots.plot_2d_separator(mlp, X_train, fill=True, alpha=.3, ax=ax)\n",
    "        mglearn.discrete_scatter(X_train[:, 0], X_train[:, 1], y_train, ax=ax)\n",
    "        \n",
    "        ax.set_title(\"n_hidden=[{}, {}]\\nalpha={:.4f}\".format(n_hidden_nodes, n_hidden_nodes, alpha));"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are many ways to control the complexity of neural networks.\n",
    " - The # of hidden layes\n",
    " - The # of hidden units in each layer\n",
    " - The regularization (alpha)\n",
    " ...\n",
    "\n",
    "The weights are randomly initialized and it affects the final result of the neural network.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-22T02:50:50.799592Z",
     "start_time": "2020-12-22T02:50:24.770711Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x576 with 8 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(2, 4, figsize=(20, 8))\n",
    "for i, ax in enumerate(axes.ravel()):\n",
    "    mlp = MLPClassifier(solver='lbfgs', random_state=i, hidden_layer_sizes=[100, 100])\n",
    "    mlp.fit(X_train, y_train)\n",
    "    mglearn.plots.plot_2d_separator(mlp, X_train, fill=True, alpha=.3, ax=ax)\n",
    "    mglearn.discrete_scatter(X_train[:, 0], X_train[:, 1], y_train, ax=ax)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-22T02:51:48.163149Z",
     "start_time": "2020-12-22T02:51:48.160648Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_breast_cancer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-22T02:51:55.181902Z",
     "start_time": "2020-12-22T02:51:55.169559Z"
    }
   },
   "outputs": [],
   "source": [
    "cancer = load_breast_cancer()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-22T02:51:55.396928Z",
     "start_time": "2020-12-22T02:51:55.393182Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cancer data per-feature maxima:\n",
      "[2.811e+01 3.928e+01 1.885e+02 2.501e+03 1.634e-01 3.454e-01 4.268e-01\n",
      " 2.012e-01 3.040e-01 9.744e-02 2.873e+00 4.885e+00 2.198e+01 5.422e+02\n",
      " 3.113e-02 1.354e-01 3.960e-01 5.279e-02 7.895e-02 2.984e-02 3.604e+01\n",
      " 4.954e+01 2.512e+02 4.254e+03 2.226e-01 1.058e+00 1.252e+00 2.910e-01\n",
      " 6.638e-01 2.075e-01]\n"
     ]
    }
   ],
   "source": [
    "print(\"Cancer data per-feature maxima:\\n{}\".format(cancer.data.max(axis=0)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-22T02:52:03.479516Z",
     "start_time": "2020-12-22T02:52:03.145498Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy on training set: 0.94\n",
      "Accuracy on test set: 0.92\n"
     ]
    }
   ],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split( cancer.data, cancer.target, random_state=0)\n",
    "mlp = MLPClassifier(random_state=42)\n",
    "mlp.fit(X_train, y_train)\n",
    "print(\"Accuracy on training set: {:.2f}\".format(mlp.score(X_train, y_train)))\n",
    "print(\"Accuracy on test set: {:.2f}\".format(mlp.score(X_test, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Okay... good but not as good as svm..."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's try feature scaling..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-22T02:53:33.636513Z",
     "start_time": "2020-12-22T02:53:32.861825Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy on training set: 1.000\n",
      "Accuracy on test set: 0.972\n"
     ]
    }
   ],
   "source": [
    "mean_on_train = X_train.mean(axis=0)\n",
    "std_on_train = X_train.std(axis=0)\n",
    "\n",
    "X_train_scaled = (X_train - mean_on_train) / std_on_train\n",
    "X_test_scaled = (X_test - mean_on_train) / std_on_train\n",
    "\n",
    "mlp = MLPClassifier(random_state=0, max_iter=1000)\n",
    "mlp.fit(X_train_scaled, y_train)\n",
    "print(\"Accuracy on training set: {:.3f}\".format( mlp.score(X_train_scaled, y_train)))\n",
    "print(\"Accuracy on test set: {:.3f}\".format(mlp.score(X_test_scaled, y_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-22T02:54:39.400037Z",
     "start_time": "2020-12-22T02:54:38.898325Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy on training set: 0.988\n",
      "Accuracy on test set: 0.972\n"
     ]
    }
   ],
   "source": [
    "mean_on_train = X_train.mean(axis=0)\n",
    "std_on_train = X_train.std(axis=0)\n",
    "\n",
    "X_train_scaled = (X_train - mean_on_train) / std_on_train\n",
    "X_test_scaled = (X_test - mean_on_train) / std_on_train\n",
    "\n",
    "mlp = MLPClassifier(alpha=1.0, random_state=0, max_iter=1000)\n",
    "mlp.fit(X_train_scaled, y_train)\n",
    "print(\"Accuracy on training set: {:.3f}\".format( mlp.score(X_train_scaled, y_train)))\n",
    "print(\"Accuracy on test set: {:.3f}\".format(mlp.score(X_test_scaled, y_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-22T02:56:24.790293Z",
     "start_time": "2020-12-22T02:56:24.246628Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20, 5))\n",
    "plt.imshow(mlp.coefs_[0], interpolation='none', cmap='viridis')\n",
    "plt.yticks(range(30), cancer.feature_names)\n",
    "plt.xlabel(\"Columns in weight matrix\")\n",
    "plt.ylabel(\"Input feature\")\n",
    "plt.colorbar();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-22T02:59:14.652888Z",
     "start_time": "2020-12-22T02:59:14.648756Z"
    }
   },
   "source": [
    "Okay.. we have good results, the model is really hard to explain and MLP in sklearn are very limited."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It should be nice to explore other (deep learning) libraries like keras, lasagna and tensor-flow."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1.1 Strengths, weaknesses, and parameters"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Neural Networks are among the state-of-the-art ML models in many applications."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "One of the main advantagens of it is that it's able to capture patterns from a huge volume of data and to build extreme complex models."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here we need to worry with computational time, volume of data, pre-processing and fine tuning."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A common way to tune the MLP is to overfit, to make sure it works on our data and then tune the parameters to improve generalization."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "    "
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:dsp] *",
   "language": "python",
   "name": "conda-env-dsp-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}