Commit aeff8076 authored by Olivier Michel's avatar Olivier Michel 💬
Browse files

Replace N3_a_Random_Forest_Regression.ipynb

parent d948c1db
......@@ -38,7 +38,7 @@
},
{
"data": {
"image/png": "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\n",
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
......@@ -91,12 +91,12 @@
"name": "stdout",
"output_type": "stream",
"text": [
"MSE = 0.0019143711219644015\n"
"MSE = 0.001900837145744446\n"
]
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAD4CAYAAADvsV2wAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO2deXxU1dnHv4csQBIEEkLYAllYAyJIWAoiIsGCpVLF4Fp5tQpWa6m1rcVS2/ryFlurllarUJeiVZEUF0SBQoqAQNSAoBKJZEHC0hAStiSQIeG8f8zcYWYyM5lJJpnt+X4++Uzm3nPvPXcm+Z3nPuc5z6O01giCIAihTzt/d0AQBEFoG0TwBUEQwgQRfEEQhDBBBF8QBCFMEMEXBEEIEyL93QF3dOvWTaekpPi7G4IgCEHDzp07j2utE53tC2jBT0lJIT8/39/dEARBCBqUUt+42icuHUEQhDBBBF8QBCFMEMEXBEEIE0TwBUEQwgQRfEEQhDBBBF8QBCFMEMEXBEEIE0TwQ5CqGhNLNxdTVWPyd1cEQQggRPBDkJz8Mhav3UdOfpm/uyIIQgAR0CttheaRnZls9yoIggBi4QcN3rhp4mOjmTcpnfjY6IDpkyAI/kcEP0hoSzeNp0IuriNBCC7EpRMktKWbxhBygHmT0gOiT4IgtBwR/CDBcNO0BZ4KeVv2SRCEliMuHaER8bHRZGcmk5NfJv55QQghRPDDDE/881U1Jh5audvqn5fJWUEIDUTwQxhnQu3JRGtOfhmbCiuYPCjRaunL5KwgBD/iww8xqmpM5OSX2Qk1mCdfq2pM1JrqmT9lAFkZSTy9oRBQzBmfYhfCaevDN9w7ttsFQQhOxMIPAWwteUPkl28vtYq7IdQ5+WUsyS0iJjqCjQXlLMktYknu/kaWu2Mcvydx/a7cPuIOEoTAQSz8EMDWkjfEvdbUwJLc/SyYPtgq1Ma+rIwkVu8+zNyJaXSMjvCJ5e4qlNPTEE9BEFofEfwQwNEFY7hvYhzE3Ni3dHMxS3KLWDB9sM9E2JXbR9xBghA4KK21v/vgkszMTJ2fn+/vboQcxRXVPPrOlwzt3Zl7PXDVGHMCrZ2qQRCElqOU2qm1znS2T3z4YYKtL31jQTnbiitZtqWkkf/e0ecuETqCEDqISydMcPTz15rqAUV2ZrLVih+dEs/P/7WH4ooawOxzN9rWmhqoqjFZrXyx/AUh+PCJ4CulXgJmAMe01sOc7FfAEuBaoBb4H631Ll9cOxxpjtg6+vkfnDrIeq6HVu5mU2EF6YmxFFfUkJ4Ya20fHxtNTHQki9fuIyY6wjo/YBwDMhkrCMGCr1w6/wCmudk/HRhg+ZkLPOej64YFLXWzuBsglm8/wKbCCsamxjMhvRtX9O/GsjsyG8Xlz5/S32rlOy7MEgQhOPCJha+13qKUSnHTZCbwijbPEOcppboopXpqrY/64vqhjmNoo214pavFU+6Ot8c8aR8VoXgl7xsWTB9MemJco0HC1sp3fFoQ944gBAdt5cPvDdiao4cs2xoJvlJqLuanAPr27dsmnQt0HEMbjdWvtm4VW3fL8u2l2A4CtgPE0s3FdsI8Z3wqAGdNFxjVL95ukZazQcY29NPAm1h7GRwEwX+0leArJ9ucxoNqrZcBy8AcltmanQpEHAXRlUAabpUJ6QkM7dWZymoTT2/4GoAluUXAxUHA2QBhCLNhvS/J3ceC6YMBWLq5mKyMJMB+kHAl5t7E2stCLEHwH20l+IcAWzXoAxxpo2sHBYawGytkwSyIrgTSVmRt28yf0p/5U/pjROAY5zbE3pnf3dW5jEVaTQm0o8XvLs5fFmIJgv9oK8FfDfxIKbUCGAucEv+9PYbQzp/SnwXTB5sFsb6e7KHdiKhL4YZhiXDuHFW151m1s4xZmX3JHp7EqrwSsoYkcfZsCqC4Y2xfs8Aqy0OV1uR8etAq9k/OHtHIlWIr2LaC7JhszdEdBDRyIQHMfSWf4ooathVXkmBzbnHnCIJ/8clKW6XUG8BVQDegHPgNEAWgtX7eEpb5DOZInlrgTq11k0tow2GlrSGCWRlJbCwoN4vh4QPU/fIRIta8R+T51kk6pi0DglLq4uDg8NoA1F/QRLQzB3PVa4hsp9BKEdlOoZTi/AWNqUFzQSnaderE+Y4xHK2DCx1j6ND1Ek6lDaTP5G/x2SnNoX6DeHpXJQ/MHMm8q/rLACAIrYC7lba+itK5pYn9GrjfF9cKNexcKIPjYOzlsHcvxMbxz0uvYXDmEMalxgPwcclxNhdWkBAbxbRhPTlQUQ3AyOQuxERHgDF4N/G680AVHxUd54r0BEb17eKyrel8A/uOnGJwz0tAawr/e4b6+gbyvzlBv/iOfCstgaJjZ6hv0ESj6R8L5WUVRJvO0bej4sSRCnq99yYd336Fb1vu9weAfr4DF2Ji2X3ZVTw7PNt8707cRTIgCIJvkZW2fsbOp/3rX8K+ffD449Rm30Jd2XkGZiZThXlgyPpBEl+tKWBTYQUfDUpkU7sKFkwfzAQvxTK1xkR+fhldMpJYajxVOBHUjsBIm/cjLOd9duVuHre4iGxdRW9YBi8jKVtkjYlX80qZ0O40qzbt5YdxJ0lsV48qL6do99dM/M87LK84yofTX7RbxWsgE7yC4FtE8P2M1X++bh089xx1N9/KP8bdSHZSD+almQXQmDitNdUzvE9nBnTvBMDwPl1cTn66Ekt3BVKawjh24YwMxqWVWyJ5zAOQ4ZbKK6m0RvjEx0Yzd8oglm4u5iVTIkljJ1pDR9dsL+WywSO4+pnHePFvL5PTYa61T1J4RRBaBxF8f2Mywe23Q04ODBtGzsy5jUTYMce9YVnb5rp3xJVYOsud70pQHZ8SnA0QT84eYTeAbCqsYFxaOemT4lz2xSjE8sj3bmXys//LnV1qSXUyADlG/wiC0DJE8P1IVY2J7c+vYEZODjz0EFULHqVi11HmT+lOdmYyxRXVLFpTwMIZGWRnJrN8eynzpwzguhG9GJdW3kioHQXamVi6W0DliLvFVwauInxsMax1I5rnuhG9ALgxMxnVvTujos+BWPSC0OqI4LcSnkw45uSXceyDrcwAePhhcr6sYEluEfOn9Ccnv4yt+4/zUdFxoIBxaQnWoiXpiXF2FrTt+byNmXeHsxW+7o51t9+w6s1oYqItf3o9e8LRo173TRAE7xHBbyU8Ed/szGQqok9woXsS7RITyc7sDJhdN4vX7mPuxFS01gxI6mT1i7uzfp1ZyC2JdHEmwE2dz3Y/YP3dSLN81nSBvJIqPi6totbUwIM9e1J/+AgvOonxFwTBt4jgtxKeuCfiY6OJrzgAwy+1vncsT2gMHAkeWL+OAu1NGmNPB4amBjJjf62pns8PnbK79oNTzRO4H5dWAXDWVM9XKo7kg4ckGkcQ2gAR/FbCI/dEQ4M55v6++1we6+nEqt3CLYtge5PG2JMnEtuVt67OZzvB7OzatsVXADacaMePTlbyyDUDuFF894LQqojg+5NXX4Vz5+DSS1028XRiNa+kspEl7zhB6w5PnkgMP7y76CDHp5SsjCS7JwfH4it7xw6h3Y4LzB3SCcSdIwitigi+H6iqMfH+P9fx/XvvpH5UJpEzZzpt44mLxTb18fA+h+1KEXozCepJW2+iaIzzuUu+Fh8bzcQrh8OfMU/c9uzpUV8FQWgeUsTcD+Tkl/Hp+1sBWPXA/1JcH8WdL39CsSVVgtHGk6pWhrCmJ8ZZ0hzvb7WC48a1vJlYzc5MvpgMzgajitepLt3MGyyROo7VvQRB8B1i4fsYZwVIjO22ESvDEuoBGH3V5dxjyS4JBbx85xigeWmE/RnH7uqJxNWTgzGgxY3sxG0AZWV220EmcAXB14jg+xjbeHOjAImx3VbIJkRWQ/fubDhwxlo4fOGMDOt5mhOT7s84dm+F2hiUpo/oCamp8PvfQ3Y22ZnJVFbX8Z99xzh88iwdIiPoGB3htoSjIAieIYLvY2yjUJxFp1h97AcOQGqq04nVYMwS6e3Thd3gtHIljB8PCxcS/9xz7D9WzcelVdbwTbAfPB0Jxs9LEPyBCL6PsY1CcdxuWwh8XmkpZGY6tcqD0a3RoqeLzEy4+2544QX4xS9YOCMDU/2XpHePs1r4ttW7HMU9GD8vQfAHIvitgCuL02oFj+wFBw9CdrbT4zxZVetPXM1TNHWMKyu8qsbEqgmzufPFl4i49FLS09J4beRI6N8fZs+GQQOtbZ2Ju+TgEQTPEMFvBVxZnFYr+NAhOH8eUlI8Os4Rf7swXM1TNHWMq3TN5tXANay8/SmWVHxEhj4D770HJ07AkiUcfHstvynS1iRy4DqBmyAIrhHBbwWatDg//tj8mm4vUp5aqv52Ybiap2jqGNtXA2M18IT0BDKnDKDH+LvNC7C0hpISGDeO4/N/xqZrfsHBqnxy7h0v4i4IzcQnNW1bi5CsaVtfb/ZZ19RAQQFERXl9Cn9b+L6kyXv5yU+48PzzzPz1Kr44g7WaliAIznFX01YWXrUVH32E6fpZ1HbvAXv2UP7QAu7852fWxVbeLDhqzgKoQKXJe5k9m3Z1dbzRq9K6gEsWZwlC8xDBbws2boSJE6nfvJl/9xzGu799llurU9lUWMGiNQWA5ytrvSXoxXHcOEhJIe75Z5l3ZZpd5a3WWlEsCKGK+PDbgmXLIDGRc18WUrrnGDu/OUlx0XG7xVatFWnib39/i2nXDh59FO66i+pX3+C1fmMDPopJEAIVEfwW4tQHrTVUVsIll3BqTwGx77xL/Q/uJr57V2Kiq/io6DiTByXy5OwR1mNaK9IkmEIWXYZ73nEH/OUvRN13Lx9+dwHqnlnMu3qg23MJgtAYEfwW0siCPnYMbr4ZNm0CoDNQr9qxatjV3EZjAV7aypWegilk0WW4Z0QEvPMOUWPH8caKR6gveRfeWgW9e/uxt4IQfIjgtxDb9MRLNxdzx/sv0HHLFvLnPMDWokpiBqSxql0Ppg9snPM+6N0tPsZtuGe/frT7qgD+9S8iHnyQ0pv+h87r33c7UIZSNJMg+AIR/BYSHxtNdmaytZTgrLffpeOECaQ9+yd2WlbN1u4+zFlTA09v+BqAJbn7geByt7QFrtJSWOnaFe65h71rt5L+3kpe27afu68Z6rK5DKiCYI9E6XiJs6gXY/HQ9d013Yq+Im/wWAC7PPXLtpZYhF5bwwtDKbyyLYm/6QY61tfxncp9bqOQXOXiF4RwRSx8L3GVyyVx3+fM/OPPAHi0vh+z8svs9huuCknz6xnu3DEfdB3ILVEdqHnhH6zueRmL1xUCzitqiWUvCBcRwfeSRmmOLaUEpx38jIiSEsqf/Tuzhk5ulOvFratCaIQxsNaa6omJjrQT/hsm9GfvrDsYs2IZPduZuDx5CAO6pDg9j/jxBeEiIvhucCYWjdIcWyzIA18UkdTxEt4ZOlmsSh9gDJi1poZGT1TxsdF8Nvdn7C6u4KavdpGZu4Ejn+Vx4aMtjURd/PiCcBERfDe4Egtnk639L5zhVM+edpa/0HwMd0xVjYkYJ/nws4b1ZOMTT3A8I4k9N9zKkF1bybFxoxk4flfFFdUsWlPAwhkZpCfGte1NCYKfEcF3g6soGme+4ehj5ZxL6M6S3P0epwwWmsbxs3YchJduLqYyshtX1p5kdv9OQOMnM9vjF60pYFNhBbb1gwUhXPBJlI5SappSqlApVaSU+qWT/VcppU4ppXZbfh71xXVbm6aiaKpqTOz44QLqLx8FZWV0H5QqUSGthBGNk5WRxILpg8nKSOLpDYVUVpsYOnk0AF0PlVJVY+KB13exeO0+y6pdexbOyGDyoES7+sGCEC602MJXSkUAzwJTgUPAp0qp1VrrAoemW7XWM1p6PX/gauJv1x+eI+v5x63vL/To4Y/uhQXOLHtjVe7jQ/ubGxUWknOuG9uKKy1HqUbnSU+ME8teCFt84dIZAxRprUsAlFIrgJmAo+AHJRcrMlUANr788+eZtPzPnI+MIqr+PAA7TR1kgrCVcHSv2Ya6XjO6N0RGcvaLvdT2HMvciWl0jI5gzvgU/3VYEAIQXwh+b8A2T+0hYKyTdt9SSu0BjgA/01rvdXYypdRcYC5A3759fdC9lmEsqpo8KNGai31jTi4z//Mm7Q8eYN73HuGZNU8QVX+ey0YPYUGyuHRaA0dffKNQ17Q0Gv6xnOFvb6J/Yhz9br0epj7gh54KQuDiCx9+4+dmcCyjtQvop7W+DPgr8I6rk2mtl2mtM7XWmYmJiT7oXsvIykiy8/l+MPcRZt95LZFvvM4X07Opufa7FCaYB6ZOaX1l5Wwr4ja3/9y5dEjuzfCIs/SpOAg//jEfz/05pzZtbfpYQQgTfCH4hwBbk7YPZiveitb6tNa62vL7B0CUUqqbD67dalTVmFi/6Dmip05h/HOL2VhQTs4n33DVmlfZ3384f399C98dPodRKV2JGjnCfFDPnv7tdIjjrvBJ1b0P8Nc/vM4/n1nFqY93UZk6kLF//xMdvnstnD7tcdEUGRiEUMYXLp1PgQFKqVTgMHAzcKttA6VUD6Bca62VUmMwDzSVjc4UQOTkl3H588+QfLiAu5XiZPISIr76iktOH6P6z39i9rWXo7pbJnIrb4Qv8qBXL7tzyCpPz/Hks3KXbM4xtfLsDz9kw0s5TP3dfFi2jOwf/tjlsY7nkXkYIWTRWrf4B7gW+BooBn5l2XYvcK/l9x8Be4E9QB4w3pPzjho1SvuLyuOndH1UlK7Lmqo1aP3cc1pPnqx1ly5a19Z6dI7nPyzS/R5eo5//sKiVexv8NOezqqyu089/WKQrq+t0ZXWdfurf+/RT/y7UldV11v2HRo7TDZ07a71jh9fnFIRgBMjXLjTVJwuvtNlN84HDtudtfn8GeMYX12or4vd9AefPs3HS9WSVFBOxYAGcPAlLl1J1IYLlGwppKhlaVkYSeSWVjE6Jb/VCJ8FOc1JF21rj2ZnJjXLu5OSX8crYuXxQ/hs6X3EFDB4MffrAunUuzykJ14RQRlbaumLbNgAeKb+ETtNuYvzqV+G22+Cee8jZUuK8MpMDGwvKreGcjcI6BTuaI7S2g4SzZGvm/VdS9ePNHPz1YwzZ+zGR69dDYSEMkmR2Qvghgu+Kzz6joV8/5t44jsGZ2fDs7627sjOTqayuo+DoGWtBbWfYVsMal1Yu4Zo+xnaQcJVszViktbj/jfxf9p3cdvMkWLNGBF8IS6QAiiuqqojo0cMuzNKI4ABIiGvPR0XH2VhQ7vIUhiClJ8ZJuGYrY3zWc8anNEpvkZWRxIT0BAo7xFOROpDz76x2G40jkTpCqCIWvgNGtMgPqk4QmRBvt8/RZ2z7KgQGzlxDGwvK2VZcybbiSvolDGHOJ+vI+fSgy8IpEqkjhCphLfjOQgGNf/abyo/TJT3Nrr2tyMvkXvBgpGE4a7pA2plLicx/l9n9OoCLRHcymAuhSlgLvuNEX1ZGErWmeuZPGcAlL9VAly7WtrZ51MU1E1wYaRiqakx8sqcPAF2PH2XepNEu28tgLoQiYe3DN4pcg2Lx2n0sWlPAktwiYqLa0e7USeja1drWyKO+aE1I5IQLeVwVm19SZHn/zTd+6pkg+I+wtvAdqypZo2mGdgOTibyqBgZaqleZc+kUSB71IMFVmGb0mQnwMnaCLyuihXAhrAXfwPYRPn1SHBwxpwJa/U0t/Sxl8ySPenDhKkzzzu+MhE6d4OBBwE36a0EIQUTwnXHyJABXjx3I5TJxF5S4qomLUtCvH3zzjZ3YG+mvDcTqF0KRsPbhu8Qi+FnjBxEfGy1x2UGMY5nKqhoTJbEJHPvya5ZvP2AV+ydnj7ATdk+zawpCMBHWFr4zK66qxkTe1q+4FqxROhKXHTrk5JfR4Xwcs8o+oX31KesiLUcrXkIzhVAkLAW/uKKaR9/5kvoLmo9Lq4CLQp6TX8bebfvMgm+J0pF//tAhOzOZtT+4i5gH1nLXe0vp8NILTttJaKYQioSlS2fRmgK2FVfycWkVkwclkpWRZHXZzO4TyQ9izC4dw8J3dAs4Ii6f4CE+Nprb7ruBdj/9KR1efhFWrvR3lwShzQhJC7+pCbeFMzIw1X/J0N6duXdSOm9v2cez678CYN5v76brhx+aG3bu7NH1xOUThPzf/5kzot56K+f/9hxFcd1JemM58Z06+LtngtBqKHO+/MAkMzNT5+fne33c0s3FLF67z+lknB3798OJEzTcMQddUszeb9/I8PdXoIzPxMPPRiI6ggvj+7qmZxSnf/0Yfbfl0vXwAd56/i1uGN4D/vQneP11aN/e310VBK9RSu3UWmc62xeSLp3szGQmD0pkU2GF6yiLEyfgiitg7Fgi9n/N0RFjuWzNGxfF3guacvkIgYXxRPbYtv8yM/V6Vix5gwvtIphe/AksWgRvvQUbNvi7m4Lgc0LSpRMfG82Ts0dYrW6nLFwIx49zdv6D7IjtRdc5t8OUiXTv3JH2e78AzJagiHjo4axOQcP4CdQvX44+Xo4C6las5B+dhljbyhOcEAqEpOCD8ygL41H+5ogKOj/3HDzwAK9cf7/Z/bPtIDtnLeKhyWmohgZeWP8lt1lW2QqhRXxstLVKliHiuZNmMuWjh7jQLgI1aRK8+y5/6nmD9ZjFa/eRV1Lp3kUoCAFOyAq+LYbQ15oaWJK7n+vW/o7OSUnw2GNkR3YEzNbexrQEvmux6M4mJEoYZgjjONE+csGP+Odlo5mRegldDpXS/vrrebJ3DVdY/gbySiqtLkIxAoRgJSQnbR0nUY1J3PlT+tPr6AFumjMNnngCfvYz33daCAqMv5GsjCQ2FpTbu2tOn4aEBHjoIXj8cbv24tYRAp3wmrQ9e5aq797A13981jpha6RBnjM+lZsKNkFEBNx+u587KvgTw+W3saC8cQqFSy6BCRNg/fpG7QFZcyEELaEn+B07knqggAdOfW51yVijaPZ9gX7mGUrHTaaqU3wTJxLCAcMYyM5Mtl9A9+1vw+7dvPJWXqOc+pJjRwhWQk/wgYipU0n5/BPi21+8vaoaEydm3cSZ2M7cdNn3Wb69VCw1wS6k1k7Mp00DYPdLK+3E3XaAEIRgIyQFn6lT4cwZ+OQT66a17+2g6zfF5H1vDkMzB3P2/AWx1AQ77MT8ssu4kNidH1QXsnX/cYorqgFZcyEEN6Ep+JMnm/Oeb9xo3TSzyhyRcXT0FWwqrKBjVDux1AQ77MS8XTvaTfs2fXduY2fBISltKYQEIRmlA8Do0dChA2zdan5/yy2weTNVX5eSs/OQRFsITfPBB/Cd71DVNZHD7+ey3RQjfzdCwBNeUToGU6dCXh6cOcOZVe/Q8K9VnLtmGvFx7ZuMtpDsl+FJo+/92mthyxbi62q45KfzefyDAnEBCj7F8W+utbUndAU/Kwvq6+HPf6bDbbfwZbcU3rjxAetud9EWEokRfhjlDht97xMnwqJF9MvbxPZ1/8tNsdX+66QQEtiKuqPWGO8fWrm7VUQ/dFfaTpgAMTHw6KNE9OrNnmdeZebkodbd7oqaSMGT8CMnv4xNhRVMSE+gstrE0xu+Zs74FLP75ic/gc6d6fnzn8MVY6GwkKpuPWUhltAsbFd5O2pNdmZyq67qDl3Bb98e1q6lJu8TVve5nBnXjLT7x3RX0UiqHYUfxj+ckX4DICY6wvx3oBTcdRdcdhlkZkJeHjk9MqUGgtAsbMXdUWviY6NZOCMDKCArI8nn1w5dwQeqRo3joeIObPq8gtO9yxolzAL7JfMgWRHDFeMfz/wYrQHV6AmvKnUgXdpFULdrD9mPXg/IU6DgPa4MSkOLum3fTM/Nu8jt+wDpUwb59No+EXyl1DRgCRABvKC1ftxhv7LsvxaoBf5Ha73LF9d2x/LtpWwqrOCK/glWsXe0ymy3AWK1hTnxsdE8ONX5P9nynUe4rktPLmzLZ4DNP63k2RF8gaFFG7e8zjUl+zg/9kmfX6PFgq+UigCeBaYCh4BPlVKrtda2gcvTgQGWn7HAc5bXVkYBMKpfvDUlLthbZe62CYI9isJufRm5f59drQQpcSn4guzMZNqZ6kh/Jg91xx0Q5/uKa76I0hkDFGmtS7TWJmAFMNOhzUzgFW0mD+iilOrpg2u7xHg0nz9lAHPGpwDOV0nabpNVlII75oxP4fzgIXSvOMxb2/Zbt0u6BcEXxMdGc486gqqpMYcEtwK+EPzegG384iHLNm/b+JSc/DKW5BYREx0hAi74hPjYaK6+cQoR+gLZ0SfstouhIPiEdevMASeTJ7fK6X0h+MrJNsflu560MTdUaq5SKl8plV9RUdHsThlWV1ZGkiyiElqEbdx0p+lTITKSzh+s9ne3hBChuKKaO1/+xJyvaccOc5aA2NhWuZYvBP8QYPss2wc40ow2AGitl2mtM7XWmYmJic3ulNt854LgBXaLY7p1g2nTaHj9DZZu2m9nSMgKbcEWT/8eFq0pYFNhBb9/93P47DNz6G8r4YsonU+BAUqpVOAwcDNwq0Ob1cCPlFIrME/WntJaH/XBtZtEFlEJLcXxb6j6xpuJW7OGz//8Ivz0Hms450Mrd7Op0PxUKpO3gqeT+Ubc/W/TNJw926qC32ILX2tdD/wIWA98BazUWu9VSt2rlLrX0uwDoAQoAv4O3NfS63qKVCoSWoqtj76qxsT886l8kZTO7z9cxuw086O3sVJ38iCphSyY8WQyv6ryNJvzi3ly9giSSyyBja0o+KGbLdMBo67tgumDxfoSmo3xd3RHxxP8btGd7L96BlF/e4Z3is+AancxHYMguKOujhMVJ6m8aipJZUXsvv+XTKwshrffhpMnoV3zbXF32TJDeqWtLeLaEbzF2YIq27+jj7/ezrjXn4P+73BPdEfWP/MG8VMH+rPLQpBQ9/05dM15k67AN30HMfHp35hTeNx/f4vEvilCN1umAxI6J3iLs6yptn9HH9/2Q3527U/IvetnxJ4/x/SDnzk9j0zmCo6czTNX4/vXrB/Sadcn5sicXr1g0aJWvW7YWPiC4C1NPRV+f9Iglkfdy+coJuWtJSb/Y6ftZCWuAPZPjF1qz7D3ulu5evmfzXND6//Dig+/ou6TcuaM79hqhtdXe+0AAB0tSURBVKkIviC4oKkkV8ZAsCR3P1NShzH8o3Vw4UKjR3JxJwpwceBvZ6rjnsrj1HbrfnHf5+X88eNjwLGLWVpbARF8QfAS+4R75jWFR4ZezvD334S9e+HSS+3aS7ptAWwG/u4XAMg5coF0S8777Mxkak31OMvS6ktE8AXBSxwt9pjoCMZ1SoY/PmxeOOMg+IIANgP/jh0AXDX5MsZZ/obcZWn1JSL4guAljhb7vEnpUFdnjrIoLfVjz4Sg4Ig5ycC13x4FbRxEEjZROoLQqrRvD717i+ALTXP4MAAnOndr8+gtEXxB8BWpqVBSYrdJQjLDG6ff/5EjEBXFygNn2zzPlwi+IPiKtLRGFr6zWH4hfHD6/R85Ar16kT26rzX1QlsZBuLDFwRfkZqKPnyYFzYUMGt8fwAqq+u4on+3VilILQQ+TkNyS0qgd+9Geb7aYq2GCL4g+Iq0NJTWvJ7zEdVE8PmhU9bsmRsLykmfFOfnDgptjSHqhgU/u18Huu7YAQsW2LVrq7UaIviC4CtSUwHIjjjOyfMX2FRYwYT0BDJT4hv9I0vh89DG9vs9UWti7iv5FFfUMOjC51x14QLMmmXXvq3WaojgC4KvGDmSU736cvvLv+erUwe5qfRrOq84RPv/+186xdonVZN0C6HNN/Pmc6Rck7Pw5+SVVFJcUcPA+PZc8a93aOjXjxdOdSK7xtTmg70IviD4ithYdG4u1dk3M/aNpZzsmUxl3QX63TUHLhsKw4ZZm0q6hdBm5GvPMxKovuVbZF0zgz6Hinh4+yoiP/2EZff8lsXrCkGpNh/swyYfviC0FUs3F7Pk3c+4Z9owEmpOcccN4+B3v4NHH/V314RWws5FFxNlzadUP+xSIjvFwY4dNERGkXf7/dyWlMXkQYk8OXtEq1j4kg9fENqQrIwk8kr6ct3IPqQnDjYvyHKIzxdCC8NFl1dSyVPfTqUrcDQugZ5ffgHAF7PmcG/iJL4zYyyTj1WzcEaGX+ZuJA5fEHyEEYmxevcRNhVWsLGg3LwjLc1jwZeFWsGF8X1lZSQxeVAimworWLv5SwCOzLrF3Cg6mj5PLeaOm6+kY3Sk/d9GGyMWviD4CMPKmz+lv10t07p+qdSvW0+dB5N0MpkbXNh+X0/OHkFOfhnfUWYxH/idyfx353a6XjYUnZAApbVcN6IXMdERfpu7EcEXBB9hOxFrK+yfR8cz+ng5b2zbzw0TBrB8eymgmDM+BcAuPFMmc4MLx+983qR0WLcfgNxjDTw0/Tc8PG0w2sOBvLXDdUXwBcFHuIqlHjJhBLwE18Se5aGVu62LsWKiIwBYvHYftaZ6YqIjyc5MFss+iHD6nVdWAnDVhCH8YlhHbrQMCrWmempNDVS5edJr7Sc8EXxB8AHuLLO4DHOe871bPmPTqV5OF2PVmhrElRMqHD8OQJfknswbkWDdHBMdyeK1+9xWtGrtJzwRfEHwAW4ts3Tz+8nffMb8WVcSc6KS7DF9rMfZFlIRV05g45HLpbLSXBuhSxe7zZ6IeWuvuBXBFwQf4PafOTER5s2jw9LnuHPFSrqcquRUr75sfWgRi4/FA+ZBQiz7wMcY2G1dcPGx0fYDQWUldO0KERF2xwZCqUsRfEHwAU3+Mz/zDPl17SnZWUC7AQO4/sv/8J1f3MklP/kdvTMmsXRzcSOrUfLtBB7ZVn+8vQvO7gmvshISElyew5+I4AtCK+Ao1lV1F9h6631wqzk659Tp01RffyNXPrmQivdWsO2ybLj/NrtBQ0I0Aw/b7JegrZOwdk94ASz4svBKEFoBx8IXOfllLMktIiY6gvjYaHK+PsXkK3/Kx/c8RELFYX5bvr2ROyg7M9kunl8IHOJjo4mJjmRJ7n5y8suIL/ySeQNjiI+Npv5YBd+ojgG5eE4sfEFoBRx9+uZ0C5WMTom3rsysNdWzncGMOH6AtD2fNSpoHQg+X8E11u94aDdINBe8Yfdu6ktK2ZGWxNHtB3hw6kA3Z2h7RPAFoRVwFOuNBeXW+PuLcfjmML0JCf0YU7IKTp2Czp390l/Be6zfcW7uxY1jx9Khro5Vw65mPIGXmFIEXxDaAFtLf1xauZ2bZkhlPbzwFOzZA1de6a8uCi5wOXmuNTz2GKxfD5GR8P778N3vcn7y1WTNvTEgXXEi+ILQBtha/LalDudNSoejMQDUbt3Gq7q3ROUEGC4nz998E377W/Pv6elwzTXw+edE9ejBvAB9UhPBFwR/06MHDBxIzMJHUKM/JeeJJ8R3H0A4XWOhNTz8MAwfDvX1cP/95u2DBvmhh54jUTqC0MY0SoGsFGzeTN3tdzD307e59dge/3ZQsMN4OrN76jp6FA4ehHvugb174b77/NdBL2iR4Cul4pVSG5RS+y2vXV20O6CU+kIptVspJSWshLDGNmTTKv6d4mn/wjLo14/2y1+WnPgBRqNB+uuvza8OFn2g1zNoqYX/SyBXaz0AyLW8d8VkrfUIV6W3BCFcsI2vt4vXb98eRo+m5rMvWLx2H8u3H/B3V8MWR+F2HKS3rNlmbjjQeXF6Y/1FoNFSH/5M4CrL78uBD4GHW3hOQQhpbCdwG/mHhw6ly6pVtD9fBwEY1hcuOE7U2n5POfllqC07GR8Vzen4JJZvKMSobxDo9QxaKvhJWuujAFrro0qp7i7aaeDfSikNLNVaL3N1QqXUXGAuQN++fVvYPUEIbBotrho6FKU1/zckkqvHpzoNCfR0m+AcTz4rR+F2HKRPR52CAQPI2XWYJblFANa0x4E84d6k4CulNgI9nOz6lRfXmaC1PmIZEDYopfZprbc4a2gZDJYBZGZmiokjhBdDhwJwY4fTEBvN0xsKWZJbxNb9Ffxu5jA2FpRTa2pgSa65qpIhLpJ3x3M8+azcrXKOj40mvvIwpkFDqDXVM3diGh2jI8jKSHKaBC+QaFLwtdZZrvYppcqVUj0t1n1P4JiLcxyxvB5TSr0NjAGcCr4ghDUDBkBUFHz5pWWDAuCjokoWrSlgU2EFcyemMnlQIlkZSdbDAt2VEEi0+LM6fRqKi9k76iqW5BYxf0p/YqIjWL37SKOBONBoqUtnNTAHeNzy+q5jA6VULNBOa33G8vs1wGMtvK4ghCZRUfCtb8Hf/gZ9+nB35wQ6jE7l3CXxXDeiF+PSzBb+psIKxqWVWxdxSd4dz2nuZ2W4gm7/dDWx9fWkzbuDBTrJmip5/pT+zJ/Sv8kyhv6kpVE6jwNTlVL7gamW9yileimlPrC0SQI+UkrtAT4B3tdar2vhdQUhZGgUyvfaa+aiKQ88QKc7buWHv/w+D47rSXpinMUq1cyfMsCphRroYYHBTE5+GX98fy/n//JX6seMZYVOIjszmTnjU1gwfTBzxqfaZdCEwPs+WmTha60rgSlOth8BrrX8XgJc1pLrCEIo47SKUkEBlJfDrl0waxZ1P7iHf9y3iMra8yzbUsL8KQMAGvmMxZfffJqaDM/OTOby5X+lS1kp6+b9wu5zdhV1FWjfh6RWEAQ/46yKUnZmMssLz0JsBnc8vJCExY9xrLIT+266y3KUdiom4stvPsbnmVdSyZOzRxAfG83y7aUsyS2i1lTP/6R14PJXnqUu+ybG/OQuFtjUIzZwdBcF2vchgi8Ifsa2ipJRyNwomALwxYBp3Ja+ll//5wXOdD7Jivv/l6uH92b17sNW146tJeqJJRmKYZze3JOzttmZyeSVVLKpsIKc/DLL56gsRyj2/el5vtVQz6vT5lDn4XUCbW5FBF8QAgBHAcrOTKbWVA8orhvRi9zUf/KtzSvo9NhvuKdjNE/P+TVL/lPM/Cn9iY+NZunmYq9cB4HmavAF3tyTs7bxsdE8OXuE9Xswyhg+0u8CA955lcx1KykfchknktNZEqSfnQi+IAQAzgTI6s+PjSb96oFUjf0lXx45xZgXnmJyteKdHpMAsy+/KdeBswHFXftgxJt7cmxbXFHNojUFLJyRYf38l24u5sjTz/GHdX+hnTYvCYr9+0u4mzQPdCRbpiAEAI71a53lZMnJL2N2/GQKp81ixOrXyH35R9xTYl7OYus6MKJCbCNEHM/nNAOkA4EWYdIUntyTq7bGGodFawoA873H7d3DH9f9hXOTruafK7dSVXWG1wZcaVebONgQC18QAgBPJvuM3xMffR1OHCPyjjuI+/H9nJx4JW9WXPT9G08KgN0ksOP5miJY3T7NmZ9YOCMDKLC8Qs6nBxn6x8c416kLMe+s4trIjuTkl1kXuwWjdQ8i+IIQkDib7LPbFtuHE8+/QKdLM2iYfi2DIrpQesU4sv/2JNB4oGjO5GGwun2aM1ClJ8bx8p1jrO9vK91B3Dd7qPnDn6BzZ3K8nCMJVETwBSEIcGa1rjyqOTpxDvcd2MqlupqEV5/lk+j2ZP/599Y2LRGnQIsw8RTHgcoji7+hwbzgLTeXhryP6VBayvmx44j96XzruWpN9QG9itYTxIcvCAGEK7+5M59+dmYy0Q/O54Gfv8jSv7zF11dOY8yLT1E188YmzxdOuMpRb/1squtg3jyYM4ezq9dQmtCHN4dezZsP/dFcnBzz4Oe4ijYYEQtfEAIIV+4IZ+6V+Nho9h+r5uPSKj4urSJmwZ9o6HQJgz/IgepqiItzupgo1HGXy95Zu64785j94ot8ctNcZvf7LnOvTGP/sWoWXpVh1z5YXVy2iOALQgDhSlRcuVcWzsjAVP8lQ3t35o6J6cSfvx3eXwl79sCECWRlJPHGJwfZVFjB8x8WkRDXPqQWWznDXS57Z+16vGYuz7Fr1v8wv0s8O785yUdFx+2S07k7TzAhgi8IAYS3opKeGMdr94y7uGHkSAA+WvlvMkaMZmNBOQcqawEoOHqGj4pK7XP2BIjw+2Llr7erjY3Puu6vpZxJ6s3sa0eRk1/GR0XHmTwoMagteVeI4AtCKNG7N2c7d+XIpu28uHI3C2dk2K3YNQqoBFrEiS9CQD09h+PgEvnZTg4PGkYX7J8OAmUw9CUi+IIQSihF5KjLueKrIv6462s2J0Xz4BVp0LEjVSZNrames6YLAbdS1Bf+cWfnsF1BC+YFVgOSOrFsSwkA84bHE1FSQk7yRLpY8ucEyiDYGojgC0KIEZV9I71++EPyn/k+PGPZmJLC6iU5LNleDsCC6YMDqh6uL/zjzs5hrKCFAs43XOCjokqiKsp5v/BN+h8EvjY/EQyeNZ2JDoNNIHwuvkYEXxBCjXvvhdGjYcsW0BrOnYNf/5qb1/2DE9c/ACg7K9iZKyRQxK6l/bBdQbtyRymnt+7gkbVvkrLvMxp69aKKKKL//hIz776p0bHButLYHSL4ghAi2InjqFEwatTFnQcP0uG5Z7m3fXte+d59dsc5c4UEith50g93g0J6bDv+eqGAwp+/ys/XryLyv0fNO158kRfSJ7F47T4WDBjMPCfnDYUwTEdE8AUhRHArjk8/DWfO0PHPT/GSaThwhV1aYCMPT6Bl0/SkH27v+y9/IW7BAkYBZaMnkvzUkzBwIFWDL6V2e6nbuYxQCMN0RARfEEIEt+LYsSPcfTe8/jq/ToPxLhYh2YZseip2TZUGbIlLyBPRtb1v47pZGUlsLCjnB6+9DqPH8Nqv/sp3s0ZQZbnX2u0HWJJbZDeXEQ5IagVBCCLcpUpoMj3w0KEAzIg40aiNkZ4ZlNM0BO5wlcrZ2/M4w5PUELb3nZNfxlPvfc7KJ19j3xN/I/LLL4j8/u3MmTnGut/8NKDt0lGHC2LhC0IQ0SLfevfu0K0b7N3baJezMovgmaXuLLGYr1xCXvnwL+vB7Z99wC2vP8ElR80DzYVuibTLzrbrq/EaTpa9gVj4ghBEOBZKcYdT6zgjAwoKXB7j+JRga6m7sradJRZz9bThbTI3436zMpJcHmf0sfCRRcQ++GNMsXGcefUN2LGDdsVF0KOHy/sLN0TwBSGI8EawnIn1uYGDzRZ+Q4NH17MdYNy5aTwdiLx19Rj3u7Gg3GXGy1pTPXMnptH1qy84ldSbzBv+yOvJo2HcOLjkkkbtwzl7qLh0BCFEsXVfGEKb3msoWadOwbe/DW+91UgQHbGN4HFX7cnTiJbmunrcZbxcklvE5EGJXCgooDKlPwuuHdKoneH2qTU1sCR3PxA6sfXeIIIvCCGKrQgbAnh55lToF2fO/37ttfDhh9ac7waOfntfxuQ3NTC4mjNoKuNl1qBu9Jt3mPO3fM9pu+XbS1mSW8TcialhOVlrIIIvCGGAnWDedRfU1cF993F6y3beiOhtDWN0rItrTMh6mnunqsbE8u2lgGLO+BSvfOVVNSYeWrmbTYUVHuXvtxscjh6E8yYihw9rvC82GlAAdIyODEvL3kAEXxBCEHfRNVU1Jt7rNYo5wL433mVxwtXklVRacs7YR90sdxKv7u7chosFICY6witxzckvY1NhBfGxUWwqrCDHkszM1f0ZgwPAvNNfmXdkZFjPZftUMmd8il30Ubgigi8IIYg7N0xOfhmLd1QwpXc6gwt3MfnOm3jg6gEM79OZWpN5MjcmOpLFa/cxf0r/Ri4QxycAxxW6RjrmpsTVceDIzky2DjzpibHWOQPb9sbTA2BtN7VfHMx/wdxoyBBrv2xfQ3HVbHMQwReEEMTd5KghrBt6ZHBb3r+ZfvJXnHqlE6MGDOA/JSdZ0y2WfvFxPJHanymjriQ+vpPLczsOLPGx0Tw4dZBHfXRWfvHJ2SOslvvGgosVpxwt+jEpXUlJiKG4oobz994H/34P/vAH6NwZ8FzgAyVJXFuhtNb+7oNLMjMzdX5+vr+7IQhBRVPuHCPi5ouVa/nOEw9z+kwt7Wpr6XruTOOTdepkjugZOtScnqFXL+jQwbr7zLl68kqOMy6tG506eGc/njlXzys7DrD3yGmuH9mLqRk9XJ5zQ8F/efuzIwxKikMpxb7/mvs6qX0NN638K/zyl7B4sVfXB1i6udicQG364JB5AlBK7dRaZzrbJxa+IIQYTbpzjH33Z8P92agaEys/Pcg1fTqwds9hzp2/QJyu55bICjptWAfr18O//uX0Wp2Aqc3sZyfgfuPNu+7POdXddYYNg1/9qll9CJQkcW2FWPiCEGJ4YuE72+fW2j1/3hzZc+SI+XfgVO151u89yreH9qRzTFSj945t1u89yt+3lvL9cX0tJ1XMHNHb2rYpTtWe593dh+yPi4uDvn1BqUbtbatdpSfGNT5hiNJqFr5SKhv4LTAEGKO1dqrOSqlpwBIgAnhBa/14S64rCIJr3Pmv3e1za+1GRVFl0iz/RgNRzBmfYn5a2HeYE6mdmDc6nRWbi+3eA3bbsrMzOJFaRqXN4qfKlBgeHD3Qo/tasbmYxcVmYT878OI1XGFb7erlO8c4bRNuPvyWunS+BG4AlrpqoJSKAJ7F/ER2CPhUKbVaa+06oYcgCK2Oo9g1NdHpGHLpmDTNccAw0h4YMfy2CdryD1SxrbgS8NzD4E0EENhXu3J3T4FQ6KWtaJHga62/AlBOHqdsGAMUaa1LLG1XADMBEXxB8CPeip2j4BpJ0xav3WeNubc9jzFAOOacj4+N5q+3Xm4dbDzFmwgggPTEOJeWve092b6GOm0xadsbsM14dAgY66qxUmouMBegb9++rpoJgtBCmhI7x2IijnHxTZ3DmcVv+0ThC4u6pS6ZcIvPb1LwlVIbgR5Odv1Ka/2uk+2NTuFkm8vnOK31MmAZmCdtPTi/IAjNwBMXjhEnb6Q7MOLgDYvem/mC1nCfOKvUFQ6++ObSpOBrrbNaeI1DgO3w3wc40sJzCoLQylgTk2UkMS7NbOEP73MYT33oBrZPCrbnddzfHLE2zlVraggrX3xzaQuXzqfAAKVUKnAYuBm4tQ2uKwhCC7C10LtmmrNmzhmfSnxstDWvvCHSTeXXcVxRa4uRybLWVG/10Xs6CLiq1CU4p0UFUJRS1yulDgHfAt5XSq23bO+llPoAQGtdD/wIWA98BazUWjeusSYIQpviTTEQx8IlTb23vUatqZ6xqfFsKqxg+fYDTs6uHF6bXyhF3DnuaWmUztvA2062HwGutXn/AfBBS64lCIJv8can7jgB29Sr7TWW5BYxIT0BgJ3fVFnr3ho4y2RpG/JZXFFtnTS2fRV/vfdIagVBCFO8CUm0de84Zq10tt/w2Rtx+NeN6GVdCOWY9tjZxK9tyOfnh07aTRrbTh6Lv947RPAFIUxpbkhiUznvjSeHlfllFFfUsGD6YNIT43hy9giWby+1LtSytc6d+eyzMpLIK6nkgasHMC4twW7yeFxaufjrm4EIviAIXtHUilfbvPaTByXa5aS3tdptJ3CduZc2FpSzqbCC4X06ExMdSdeYiwOUkTZZ8A4RfEEQvMJY8eq4MMs2RcOTs0c4jbKxHQxsXTvu0jIAEnLpI0TwBUFoFo4Ls+CiIDvz6RvibwwGWRlJdqGdrtIyZGcmS8iljxDBFwShSZz52B0XZrkSZGdVseZNSremYza2217D1uIPt/QHrYkIviAITeLMx24rxO586q6igRyzbTpeQ0Te94jgC4LQJJ6EcNqGa84Zn2J9EnBloTtm2wy3zJX+QARfEIQm8cSt0lS4pjPEddO2iOALguATvC1QAuGXntjfiOALguATvC1QIrQ9LUqeJgiCIAQPIviCIAhhggi+IAhCmCCCLwiCECaI4AuCIIQJIviCIAhhggi+IAhCmKC01v7ug0uUUhXAN808vBtw3IfdaWuCvf8Q/PcQ7P2H4L+HYO8/tP099NNaJzrbEdCC3xKUUvla60x/96O5BHv/IfjvIdj7D8F/D8HefwisexCXjiAIQpgggi8IghAmhLLgL/N3B1pIsPcfgv8egr3/EPz3EOz9hwC6h5D14QuCIAj2hLKFLwiCINgggi8IghAmhJzgK6WmKaUKlVJFSqlf+rs/3qKUekkpdUwp9aW/+9IclFLJSqlNSqmvlFJ7lVLz/d0nb1FKdVBKfaKU2mO5h9/5u0/NQSkVoZT6TCm1xt99aQ5KqQNKqS+UUruVUvn+7o+3KKW6KKX+pZTaZ/l/+Jbf+xRKPnylVATwNTAVOAR8CtyitS7wa8e8QCl1JVANvKK1Hubv/niLUqon0FNrvUsp1QnYCXwvyL4DBcRqrauVUlHAR8B8rXWen7vmFUqpnwKZwCVa6xn+7o+3KKUOAJla66BceKWUWg5s1Vq/oJSKBmK01if92adQs/DHAEVa6xKttQlYAcz0c5+8Qmu9Bajydz+ai9b6qNZ6l+X3M8BXQG//9so7tJlqy9soy09QWUZKqT7Ad4AX/N2XcEQpdQlwJfAigNba5G+xh9AT/N5Amc37QwSZ2IQSSqkUYCTwsX974j0Wd8hu4BiwQWsdbPfwZ+AXwAV/d6QFaODfSqmdSqm5/u6Ml6QBFcDLFrfaC0qpWH93KtQEXznZFlSWWaiglIoDVgE/0Vqf9nd/vEVr3aC1HgH0AcYopYLGvaaUmgEc01rv9HdfWsgErfXlwHTgfou7M1iIBC4HntNajwRqAL/PKYaa4B8Ckm3e9wGO+KkvYYvF770KeE1r/Za/+9MSLI/hHwLT/NwVb5gAXGfxga8ArlZK/dO/XfIerfURy+sx4G3MLttg4RBwyObJ8F+YBwC/EmqC/ykwQCmVapkkuRlY7ec+hRWWCc8Xga+01k/5uz/NQSmVqJTqYvm9I5AF7PNvrzxHa71Aa91Ha52C+X/gP1rr2/3cLa9QSsVaJv2xuEKuAYImck1r/V+gTCk1yLJpCuD3wIVIf3fAl2it65VSPwLWAxHAS1rrvX7ullcopd4ArgK6KaUOAb/RWr/o3155xQTg+8AXFh84wCNa6w/82Cdv6Qkst0R9tQNWaq2DMrQxiEkC3jbbD0QCr2ut1/m3S17zAPCaxfgsAe70c39CKyxTEARBcE2ouXQEQRAEF4jgC4IghAki+IIgCGGCCL4gCEKYIIIvCIIQJojgC4IghAki+IIgCGHC/wO3EyrPpj2UCgAAAABJRU5ErkJggg==\n",
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
......@@ -134,12 +134,11 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"## Exercize\n",
"## Exercize 11\n",
"\n",
"- Change the parameter max_depth in the code above. \n",
"- Compare the behaviour of the random forest regressor with the behaviour of the tree regressor of notebook N2_a_Regression_tree, when max_depth is changed.\n",
"\n",
"Explain your findings. "
"- Explain your findings. "
]
},
{
......@@ -155,19 +154,19 @@
},
{
"cell_type": "code",
"execution_count": 3,
"execution_count": 4,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"MSE = 0.004553804568572212\n"
"MSE = 0.000706083859664731\n"
]
},
{
"data": {
"image/png": "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\n",
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
......@@ -181,7 +180,7 @@
"source": [
"from sklearn.ensemble import ExtraTreesRegressor\n",
"\n",
"clf2=ExtraTreesRegressor(n_estimators=1000, criterion='mse', max_depth=5,\\\n",
"clf2=ExtraTreesRegressor(n_estimators=1000, criterion='mse', max_depth=8,\\\n",
" random_state=None)\n",
"clf2=clf2.fit(X,y.ravel())\n",
"\n",
......@@ -203,7 +202,7 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"## Exercize\n",
"## Exercize 12\n",
"- Study the Extremely Randomized Regressor behaviour for max_depth parameter values (change it in the code above) ranging from 1 to 6. \n",
"- Explain the green curve observed for max_depth=1\n",
"- Propose a method for setting the optimal value of max_depth parameter. Implement it (hint: look at notebook N2_a_regression_tree)\n",
......@@ -241,7 +240,7 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.8.3"
"version": "3.8.2"
}
},
"nbformat": 4,
......
%% Cell type:markdown id: tags:
This notebook can be run on mybinder: [![Binder](https://mybinder.org/badge_logo.svg)](https://mybinder.org/v2/git/https%3A%2F%2Fgricad-gitlab.univ-grenoble-alpes.fr%2Fchatelaf%2Fml-sicom3a/master?urlpath=lab/tree/notebooks/8_Trees_Boosting/N3_a_Random_Forest_Regression.ipynb)
%% Cell type:markdown id: tags:
## RANDOM FORESTS regressors
%% Cell type:markdown id: tags:
### Consider first the same example as in notebook `N2_Regression_tree.ipynb`
This is a regression problem. Rather than evaluating the optimal tree structure of a single tree, random forest is considered.
See
https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.RandomForestRegressor.html
%% Cell type:code id: tags:
``` python
# Create a new realization od the signal
import numpy as np
%matplotlib inline
import matplotlib.pyplot as plt
noise_std=.1
X=np.arange(0,2*np.pi,.01)[:,np.newaxis]
nx=np.random.randn(X.shape[0],1)*noise_std
y=np.sin(X)+np.random.randn(X.shape[0],1)*noise_std
print("The number of point in the set is {}".format(len(X)))
plt.figure()
plt.scatter(X,y,s=1)
plt.xlabel('X')
plt.ylabel('y');
```
%%%% Output: stream
The number of point in the set is 629
%%%% Output: display_data
[Hidden Image Output]
%% Cell type:markdown id: tags:
In **random forest**, each tree in the ensemble is built from a sample drawn with replacement (i.e., a bootstrap sample) from the training set.
Furthermore, when splitting each node during the construction of a tree, the best split is found either from all input features or a random subset of size max_features.
(Note that on thour example, we only deal with a single feature.)
The purpose of these two sources of randomness is to decrease the variance of the forest estimator. Indeed, individual decision trees typically exhibit high variance and tend to overfit. The injected randomness in forests yield decision trees with somewhat decoupled prediction errors. By taking an average of those predictions, some errors can cancel out. Random forests achieve a reduced variance by combining diverse trees, sometimes at the cost of a slight increase in bias. In practice the variance reduction is often significant hence yielding an overall better model.
*(extracted from https://scikit-learn.org/stable/modules/ensemble.html#forest)*
%% Cell type:code id: tags:
``` python
# random forest estimator :
from sklearn.ensemble import RandomForestRegressor
clf=RandomForestRegressor(n_estimators=100, \
max_depth=5,\
random_state=None, \
criterion='mse')
clf=clf.fit(X,y.ravel())
Ntest=300
XX = np.linspace(X.min(),X.max(),Ntest)
y_ref=np.sin(XX)
XX = XX.reshape(len(XX),1)
yp = clf.predict(XX)
plt.scatter(X,y,s=1)
plt.plot(XX,yp, color='red')
error=np.square(y_ref-yp).sum()/Ntest
print('MSE = {}'.format(error))
```
%%%% Output: stream
MSE = 0.0019143711219644015
MSE = 0.001900837145744446
%%%% Output: display_data
[Hidden Image Output]
%% Cell type:markdown id: tags:
## Exercize
## Exercize 11
- Change the parameter max_depth in the code above.
- Compare the behaviour of the random forest regressor with the behaviour of the tree regressor of notebook N2_a_Regression_tree, when max_depth is changed.
Explain your findings.
- Explain your findings.
%% Cell type:markdown id: tags:
In **extremely randomized trees**, randomness goes one step further in the way splits are computed.
As in random forests, a random subset of candidate features is used (*again, here we have a single feature, so this does not apply*) but instead of looking for the most discriminative thresholds, **thresholds are drawn at random** for each candidate feature and the best of these randomly-generated thresholds is picked as the splitting rule.
This usually allows to reduce the variance of the model a bit more, at the expense of a slightly greater increase in bias.
%% Cell type:code id: tags:
``` python
from sklearn.ensemble import ExtraTreesRegressor
clf2=ExtraTreesRegressor(n_estimators=1000, criterion='mse', max_depth=5,\
clf2=ExtraTreesRegressor(n_estimators=1000, criterion='mse', max_depth=8,\
random_state=None)
clf2=clf2.fit(X,y.ravel())
Ntest=300
XX = np.linspace(X.min(),X.max(),Ntest)
y_ref=np.sin(XX)
XX = XX.reshape(len(XX),1)
yp2=clf2.predict(XX)
plt.scatter(X,y,s=1)
plt.plot(XX,yp2, color='green')
# MSE evaluation
error=np.square(y_ref-yp2).sum()/Ntest
print('MSE = {}'.format(error))
```
%%%% Output: stream
MSE = 0.004553804568572212
MSE = 0.000706083859664731
%%%% Output: display_data
[Hidden Image Output]
%% Cell type:markdown id: tags:
## Exercize
## Exercize 12
- Study the Extremely Randomized Regressor behaviour for max_depth parameter values (change it in the code above) ranging from 1 to 6.
- Explain the green curve observed for max_depth=1
- Propose a method for setting the optimal value of max_depth parameter. Implement it (hint: look at notebook N2_a_regression_tree)
- Why does the Extremely Randomized Regressor exhibit good performances, although splits are chosen at random?
%% Cell type:code id: tags:
``` python
```
%% Cell type:code id: tags:
``` python
```
......
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment