From 31080f034f16ba46d31ad271ba8252143e2f7f40 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?F=C3=A9lix=20MARQUET?= <72651575+BreizhHardware@users.noreply.github.com> Date: Mon, 29 Sep 2025 14:39:18 +0200 Subject: [PATCH] Obisidian vault auto-backup: 29-09-2025 14:39:18 on . 1 files edited --- ISEN/IA/CIPA4/TP/TP2/tp2_IA.ipynb | 143 ++++++++++++++++++++++++++++-- 1 file changed, 137 insertions(+), 6 deletions(-) diff --git a/ISEN/IA/CIPA4/TP/TP2/tp2_IA.ipynb b/ISEN/IA/CIPA4/TP/TP2/tp2_IA.ipynb index 1c7d196..9e255da 100644 --- a/ISEN/IA/CIPA4/TP/TP2/tp2_IA.ipynb +++ b/ISEN/IA/CIPA4/TP/TP2/tp2_IA.ipynb @@ -190,7 +190,7 @@ }, { "cell_type": "code", - "execution_count": 15, + "execution_count": 20, "id": "d333dc4d", "metadata": {}, "outputs": [ @@ -208,9 +208,9 @@ }, { "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeEAAAH5CAYAAAC2z4b4AAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjYsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvq6yFwwAAAAlwSFlzAAAPYQAAD2EBqD+naQAAHwtJREFUeJzt3QeUHWX5P/BJoYZAQKQXBREJvQkC0gVEAgiKiEoRJSgQSxDRcMACIvUoRRGQokiREhQLSARUijTpQs4hghTpMQm9zv888/vPnrt37252M5EnJJ/POUuWe+/MvHdm7nzfMu/dQWVZlgUA8LYb/PZvEgAIQhgAkghhAEgihAEgiRAGgCRCGACSCGEASCKEASCJEJ5DXHvttcU888xT7LffftlFAeD/E8Kzkeuuu64YNGhQ8Z3vfKfHc1tssUXxi1/8ojjzzDOLc889t5gVbL755lV5efs9/PDD1b7fe++9Z6nzFOY0QrjlgtT6M/fccxfLLrtsscceexR33313MTv41Kc+VZxwwgnFl770peKuu+4q5jRxXCP4eefLrEQMRFQ0opxR8ZiR5eLn4IMP7vV13/zmN7te116pqSu58847b/Hvf/+74/If+MAHelSE60rS/vvv3+P1N9xwQ/HJT36yWHrppatr5MILL1ytI66TdeW+0/W0r5/3vOc9xZxsaHYBZiUrrrhi8dnPfrb6/YUXXij+/ve/FxdccEFx2WWXFX/+85+LjTfeuJiVffCDHyzuv//+YtFFF+31NV/72teKhRZaqKpYrLnmmkWmaJm/9NJLqWWAWdnQoUOL8847r/jhD39Y/d7qjTfeqD5D8Xj83ptXX321OOyww4pf/vKXjcpyzjnnFJ///Oer7W2//fbFSiutVIXoxIkTiz/84Q/FX//612KvvfYqRowYURxxxBE9lv/ud79bXXu++tWvdnt8xIgRxZxMCLd43/ve16M2GSfvUUcdVYwbN27Atdm32/zzz1/VSqcnPkizguWWWy67CDBL++hHP1pcccUVxe9+97ti55137vZcBN+TTz5Z7LjjjsVvf/vbPhsX559/fvGNb3yjWGONNWaoHFFZHjNmTDF8+PDixhtvLFZdddVuz7/++utd18cI1U5DDRHCvT03J9MdPR0HHXRQ9e+tt97ao1vz8ccfL/bcc89iiSWWKAYPHtwtpKNWOGrUqKpVGjdERa0xAr295dc6PhYnd4zdxon+7ne/u/jyl79cvPzyy9Xrfv/73xcf+tCHimHDhhWLL754ccghh/So/fY11vb0009XreCoaER5FllkkeoD/pe//KXHa6N7KH6mTJlSHHjggVW3fNR+oyZci5b07rvvXiy55JJVt9Tyyy9f7avnnnuu0ZhwbCMei3//9Kc/FRtttFFVuXjXu95V1bI7rT9uOov3stRSS1XvLfbPhz/84eL000/vtl9CvN/WrrD6PU2dOrU45phjis0226xaT7yn+DeO76RJk/rsZowL3FprrVXMN9981f74yle+0nXc2sV5ERfTKGOUNfbtLrvsUlx//fXdXhd/3Oyss86qel8WXHDBah+st9561WMD8eabb1bvK457dEvGv0cffXTx1ltvDbjLvj4v+iv2waGHHlq9x9j2aqutVpxxxhm9vn78+PHFpz/96aqM8X6j1RTH8dJLL+32ujhm733ve6vfowu09XjWn8H//Oc/VWtsww03LBZbbLFqX0fZ4zMVn4V2cfwPP/zwYuTIkcUCCyxQ7fMoR5xz7V25/T02sR8jeEJ8rmek+zXOjQiuTsc9Hovu4I9//ON9ruPII4+sjnd0Xc+oe++9t3j++eer99EewGGuueYqPvKRj8zw+udkWsL91B4WEQYRihFmEUavvPJK9YEMP/3pT4sDDjig+vBEEMdF4Lbbbqta1BEY8RMX+VY333xzdbHcdttti9GjR1evifVMmzatWkeMfe20007VNiOQjzvuuOpiEReO6YkQiQtCXHx22GGHYtdddy0mT55cXfS23HLLqpsqxnTau7DiueiWj5p2hHAER4ha92677VZVPKJMcZH95z//WZxyyinFVVddVb2XuDg0EduI9xnvPYI4wiu63uK9tAZW/ZrY11GWCMFnnnmmGvOO9xV3g8dFLy7IcUGMykLrOGKEZ4hu/NiXcZGJi1pUdh544IEqYGMb//jHP6pl28V7vvLKK6ttx/6K30866aTi2WefLX71q191e+2Pf/zjqiIUYR3biJ6AqMjF+7nkkkuKTTbZpOsi/5nPfKYaConKWxybOF+uvvrqYt9996329fHHH9+v/RjvPy7WEVpxTsZ5euKJJ1YVvv+luOjHeTNhwoRi9dVXr95DfGbi/cc+7uRb3/pW9T5jP9THMc6DT3ziE9U+rSvEccyiohP7M4ZUWluIdcDF+RL3P2y11VbFBhtsUIXEHXfcUX2m4hyN4xkhX+/v+NzFeRvBut1221XndoRvbP9zn/tc17EfyLGpz7Oo+EWY12UbSPdrVF6iYhKVl6eeeqrrMxi/x3kZxzde05f47Ecl9Y9//GN1Xelt//clKsHhX//6V1WxGzJkyIDXQS/i7wnP6R566KH4m8rltttu2+O5ww8/vHpuiy226Hos/j9+9tlnn/KNN97o9vr77ruvHDp0aLnmmmuWzz77bLfnjj766Gq5448/vuuxa6+9tmt9l19+edfjr732WrnGGmuUgwYNKhdddNHylltu6Xpu2rRp5WKLLVYussgi1eva13XEEUd02+5GG21ULr744uXEiRO7PT558uTy/e9/f7WeF198sevx5Zdfvmt/vPTSS92Wife04IILlksvvXT58MMPd3vuggsuqJY78MADy/7YbLPNqte3Ovvss6vHYh9ef/31XY/Hft58882r52666aaux3fZZZfqsTvvvLPH+tv3f7wuttnJlClTyueee67H49dcc005ePDg8gtf+EK3x2Mfx/oWWmih8oEHHuh6PPZX7NNY5vHHH+96PMoXjy211FLV+dbqrbfe6vba008/vev8aj2+r776ajlq1Kjqudtuu62cnvp8iHPxhRde6Hr8scceq86peG6vvfbq9z6K8yJ++qM+jtttt123z8jdd99dzj333B3P00mTJvVYz/PPP1+uvvrq1X5uPUfrz2x7+WtPPfVUtWy7c889t1ruyCOP7FameGznnXfu8fpXXnml23oGemzq8ySOxUDUy8VnKtYXvx977LFdz8fv8djtt9/e9blr35/15+uJJ54o77rrrur8W3/99avzrbbyyiv3+AzW583o0aO7Hotl1l133erxTTbZpDzjjDPKe+65p8f1ry+xbH/PnzmJ7ugWDz74YNXNGD8xfrLpppsW3/ve96qaZrRiW0Xt99hjj+1RI/zZz35WdROffPLJXbXHWnQhRzdz1KLbRe00WlO1qLlHCyDO3Wjprb/++l3PRXd1tGijNfvYY4/1+Z6i9h+tnn322adqtUcLrf6JGm3cMR3r6dQyivcXrbZW0RqN1nl0aba3DKNHYJ111ikuvPDCoqloYbTeCBf7OVoT7UMDtfZyhvb935doFcX+aVd3v0WLrpNoka288srdyhEtl2gJ3n777d3Oi3gsugbbuyOjlyW6vltb19ESP/XUU6vzoPWcq8/DTudQuzhWIVr4sb5a3Nka5f5fqrcd5W39jESrOFqWnaywwgo9HovenmhRRndxp+Pem+h9imXbxbajx6rT8ex0DkU3dut6ZtaxGYh11123Gss9++yzux6L36MXID5v/RHLx02nsQ8vvvjiAZchztHorYnPZPTcfPGLX6yOZezLrbfeuhoiiOsJA6c7ukV0ddZjOPEBi66fCIMY14oTrlV073W6CznuqA7R5RV3VLeL9UY3Z7u6W7RVdMlN77kY+6rHxzqpyxN3V8ZPb5544olu/x8Vj/b33Lq+6LrrNFYa3Z11yPd1l3Z/LjztlllmmerfGKtuDf64ez3G/uJYRfdjjCPOyLZjPPFHP/pR9d6i/K1j7u3DBwMt5y233FL9u8022/RZhrhn4J577qlCOYYn2sUNMKHTOdSunoYW+6Ndp8dmpth2hFWnkIht//znP+/xeAyXxDka3abRFdw+rh7n+kDEeRGVn+h6/u9//9stJFrXtcoqq1QhFeEZldro3o4u3PjcRbf0/+LYDFTcTBl3Fd90001dwyfRHT8Q3//+94uLLrqoujclxprb77aenqg8RgDfeeedVSUmhthiylJc5+InKl5x7KLiQv8J4RYxLhRjev1Rj820i1ZlaG85T089ntyq/pD09Vz9we9NXZ6vf/3r1bhQb9pvtoiWRKcv0qjXFy2Bvrz44ouNQriv99x6MY05i5dffnk1znnaaadV5YpyRws2xgQ7VWA6idZB9ApEqyfOg7jgxA039c1bvc2z7G85oyUX66orT72JsIjejxgrriuEve3f6YltRoh0Og69nb8zS2w77hXopNO247yK3p5HHnmkam1F6yrGTqMVHRf93/zmN9V9Cv0Vxz7m10bPU1R8omJUt3SjotW6rjhe11xzTdUDFjeBjR07tno8lo0bE2NmRJRjZh6bgYpWbPSk1TdoRaUwxqYHIu5BiPsC4rMSNy3GTWozIj5TrZ+rqLxG+WK8+Sc/+Uk17k//CeEZ1Ns3PdUX5eiyjW7jbHV5olUSF7aZ9f6iRRB3u84Kohs/fuLuzaiZRwsoWlpxg020SvpzI0xcgKP1H13IccNNq5nRvR5liAt49DhEd3Bv6v0bLexoaTQRXezRBR6t+giUVnFjT2/Hvbc5pxGs9c1M/dl23FjVSadtx/GKAI7WWrTUWkXrOEK4v6L8sZ6o8ESAR4WyFscghlk6DV3EEFLcABbnTIRy/H/c0Be9V3HT2Mw8NgMV5YtzPFqyIVrrAxluqUWFIoI8htnizv+ZIXoNYn9Haz32mxAeGGPCM1ncidnabTurlCfmGf7fvREzZ311t9isJCo9EbxRy49xxLjYR9dyLVqFvY1bRdd6dEu2B3CEZtwROjO+SCXEtKvpvYcoR3Q3tnZnz4j6y1j+9re/9Xiu02Mh7mqPll67+BakgZQnth0twugK7s+266GN1vsi+np9Pc7c6XhGpSMqDDGToDWAQ4Rnb9PH6kpI7P9oMcYdz6Gegzsjx6avcg5UhFxUNONnRuf6x30PMVUpPhvRWzCzdBp/p3+E8EwWXTzRvRXTKaJm3y4+vHGz1NslLv4RnLHNTmPC0Z0d38jTX3GDV1yMokZ933339Xg+xs3ezgpITEXpdIGr54K2Tt+IC1BvN7LFTWZxY15rKy3Gt+MrPqfX5d8f8RWAcUGOVl6neaetY5TxpQixH+Pml05dmw899FAVitNT3wAVrZ7W9UTI9jaeGF3Cse7W+eOvvfZaNZwxEPW24zxpPT7Rg9Lpm5vqm/za50vHFLH4UopOlYUIzEcffbTHcxG80fUcFYDWefnRnVxPc2oV77fT/qzPhdZzaKDHpr7Zr1M5Byq61WPoJX6azMmNm/KiNyZCuL+ViXhfcVNaVADaxf6oz6d6mh39pzt6Josu2hgXiYt33DUbX+8W31gTJ2+0qOLiFq20GL98u8QNJzFG+u1vf7sa+4zfY0wpPlhxQ0WUrf66zump7+6Osdho7UTLM76lK8bY6ot3zOvt79h6U3FRjACLD3+M48aFOS7kcSNU3KzVelGIeby//vWvq668tddeuwrFmMsaN+XExTl+4vG4Kz26NKMlFAEZ77Ppd23HTW4xFhnljfH3KEMET3zjUVQkPvaxj1XPh5gnHhWZ+CKK6F6PYYS4GShCIbpKo3Uf4TS9L32I4xyVpriTNrYfc5PjOEWXZuyb6B1pF2EbrfU4b+Mu7xgXj/0Q3enTG89uFXeyRxnjPIh9GvcjxLhvnDsRJu3bjtCOm53iGMTYYuyb2OdxfsZNRDHE0N7yigpD7LtYNnowoqejntMbleEImTh2MbsghofipqF4rvVO9BBd1rGNqLDGl3XEl+9ERSXCLtbZ2r060GNTf0lHfPai0hrd9LEvY6x5oOp5+U1FBSWGX6Ii0SlUO4mehTg2MWskPlNxnYv1xH6K+coxBzy66TtVcpiO7DlSs/o84U76mktZi3m9u+++ezUvdK655qrmZa6zzjrloYceWt5///3TndvbOtcy/m3Xaf5hX+uKOcGHHXZYudpqq5XzzTdfucACC5QrrbRSuccee5SXXXbZgOeDxtzYfffdt3pdzPtceOGFq/mcY8aM6TaneUbnCXd6z53e34UXXljutttu5YorrljOP//81XzSmBd7zDHH9JgnGvMl47VxLGLOZOt2Yh7kaaedVq666qrlvPPOWy6xxBLV+3v66ac7lrOv+Z/Tew877LBDNTc79tsyyyxT7rrrruUNN9zQ47UXXXRRufXWW1f7Ns6hmJsdc6VPOOGE8plnnunXPo55nDE/fYUVVqi2F//+4Ac/KB988MFe59lefPHF1bGM18d+OOigg6p9OZB5wiHm9R5yyCFVueeZZ55y5MiR1Tzb3s7TmEu9zTbbVO93+PDh1X6fMGFCr/sz5r1vv/325YgRI6r59K3HI+bwHnXUUdU5HttebrnlyrFjx3Z8H48++mj1udxwww2r+ffxvuP1MQe9dU76jB6bc845p9qfUY7+zpVtnSc8Pf2ZJ9zpvFhllVW6vqNgevOEY770pZdeWu63337V5ys+Q0OGDKnef8wbPvHEE8uXX365z3KaJ9zZoPjP9IIaAJj5jAkDQBIhDABJhDAAJBHCAJBECANAEiEMAEmEMAAkEcIAkEQIA0ASIQwASYQwACQRwgCQRAgDQBIhDABJhDAAJBHCAJBECANAEiEMAEmEMAAkEcIAkEQIA0ASIQwASYQwACQRwgCQRAgDQBIhDABJhDAAJBHCAJBECANAEiEMAEmEMAAkEcIAkEQIA0ASIQwASYQwACQRwgCQRAgDQBIhDABJhDAAJBHCAJBECANAEiEMAEmEMAAkEcIAkEQIA0ASIQwASYQwACQRwgCQRAgDQBIhDABJhDAAJBHCAJBECANAEiEMAEmEMAAkEcIAkEQIA0ASIQwASYQwACQRwgCQRAgDQBIhDABJhDAAJBHCAJBECANAEiEMAEmEMAAkEcIAkEQIA0ASIQwASYQwACQRwgCQRAgDQBIhDABJhDAAJBHCAJBECANAEiEMAEmEMAAkEcIAkEQIA0ASIQwASYQwACQRwgCQRAgDQBIhDABJhDAAJBHCAJBECANAkqFZG4bZwZtvvtlo+alTpxazg1NOOaXR8i+99FLjMkycOLHR8qeeemrjMhx88MGNlr/gggsal2HeeedttPyhhx7auAxHHHFE43XMKbSEASCJEAaAJEIYAJIIYQBIIoQBIIkQBoAkQhgAkghhAEgihAEgiRAGgCRCGACSCGEASCKEASCJEAaAJEIYAJIIYQBIMjRrw7yzPfLII42Wf+211xqX4cYbb2y0/PXXX9+4DFOmTGm0/CWXXNK4DPyfZZddttHyBx10UOMyjB8/vtHyw4cPb1yGNddcs9Hym222WeMy0H9awgCQRAgDQBIhDABJhDAAJBHCAJBECANAEiEMAEmEMAAkEcIAkEQIA0ASIQwASYQwACQRwgCQRAgDQBIhDABJBpVlWWZtnBx33HFH43VsueWWjZafOnVq4zIw+xgyZEjjdZx11lmNlh82bFiRbamllmq8joUXXrjR8iuvvHLjMtB/WsIAkEQIA0ASIQwASYQwACQRwgCQRAgDQBIhDABJhDAAJBHCAJBECANAEiEMAEmEMAAkEcIAkEQIA0ASIQwASYQwACQZVJZlmbVxckyePLnxOjbYYINGy0+aNKlxGZg5x6LpH4EP1157baPl55577sZlmDp1auN1wNtNSxgAkghhAEgihAEgiRAGgCRCGACSCGEASCKEASCJEAaAJEIYAJIIYQBIIoQBIIkQBoAkQhgAkghhAEgihAEgydCsDZNnkUUWabyO4447rtHyV1xxReMyrL322o2WHzNmTJFtrbXWaryOCRMmNFp+2LBhjctw7733Nlr+pJNOalwGeCfSEgaAJEIYAJIIYQBIIoQBIIkQBoAkQhgAkghhAEgihAEgiRAGgCRCGACSCGEASCKEASCJEAaAJEIYAJIIYQBIIoQBIMmgsizLrI0z55o2bVrjdQwfPrzR8qNHj25chjPPPLPR8uedd17jMuyxxx6N1wHk0BIGgCRCGACSCGEASCKEASCJEAaAJEIYAJIIYQBIIoQBIIkQBoAkQhgAkghhAEgihAEgiRAGgCRCGACSCGEASDI0a8PM2RZccMHsIhQLLbRQdhEa/z3isPvuuzdafvBgdXHI4tMHAEmEMAAkEcIAkEQIA0ASIQwASYQwACQRwgCQRAgDQBIhDABJhDAAJBHCAJBECANAEiEMAEmEMAAkEcIAkEQIA0CSQWVZllkbh0wvvvhi43WMGjWq0fLXXXdd4zJceeWVjZbfZpttGpcBmDFawgCQRAgDQBIhDABJhDAAJBHCAJBECANAEiEMAEmEMAAkEcIAkEQIA0ASIQwASYQwACQRwgCQRAgDQBIhDABJ/D1haGDSpEmNll9nnXUal2HEiBGNlt9iiy0al2G99dZrtPwBBxzQuAyDBg1qvA54u2kJA0ASIQwASYQwACQRwgCQRAgDQBIhDABJhDAAJBHCAJBECANAEiEMAEmEMAAkEcIAkEQIA0ASIQwASYQwACQRwgCQZFBZlmXWxmFON378+Mbr2GeffRotP23atCLb0Ucf3Xgde+65Z6Pll1xyycZlgIHSEgaAJEIYAJIIYQBIIoQBIIkQBoAkQhgAkghhAEgihAEgiRAGgCRCGACSCGEASCKEASCJEAaAJEIYAJIIYQBI4u8JwzvcPffc02j5sWPHNi7DhAkTimz7779/o+XHjRvXuAxLL71043UwZ9ESBoAkQhgAkghhAEgihAEgiRAGgCRCGACSCGEASCKEASCJEAaAJEIYAJIIYQBIIoQBIIkQBoAkQhgAkghhAEgihAEgyaCyLMusjQP5pkyZ0ngdV1xxRaPl995778ZlaHop22qrrRqX4eqrr268DuYsWsIAkEQIA0ASIQwASYQwACQRwgCQRAgDQBIhDABJhDAAJBHCAJBECANAEiEMAEmEMAAkEcIAkEQIA0ASIQwASfw9YSDdPPPM03gdr7/+eqPl55prrsZluOqqqxotv/nmmzcuA+8sWsIAkEQIA0ASIQwASYQwACQRwgCQRAgDQBIhDABJhDAAJBHCAJBECANAEiEMAEmEMAAkEcIAkEQIA0ASIQwASYQwACQZmrVhYOa4++67Gy1/ySWXNC7Drbfe2mj5119/vcg2cuTIxuvYdNNNZ0pZmHNoCQNAEiEMAEmEMAAkEcIAkEQIA0ASIQwASYQwACQRwgCQRAgDQBIhDABJhDAAJBHCAJBECANAEiEMAEmEMAAk8feEoYGJEyc2Wv7kk09uXIbLLrus0fJPPvlkMTsYOrTZ5WzJJZdsXIbBg7VrGBhnDAAkEcIAkEQIA0ASIQwASYQwACQRwgCQRAgDQBIhDABJhDAAJBHCAJBECANAEiEMAEmEMAAkEcIAkEQIA0ASIQwASZr9FWxI1PSP0Z9//vmNy3DKKac0Wv7hhx9uXIbZwfrrr994HePGjWu0/I477ti4DDBQWsIAkEQIA0ASIQwASYQwACQRwgCQRAgDQBIhDABJhDAAJBHCAJBECANAEiEMAEmEMAAkEcIAkEQIA0ASIQwASfw9YWbIU0891Wj5++67r3EZDjzwwEbLP/DAA43LMDvYYIMNGq/jkEMOabT8Tjvt1LgMgwdrU/DO46wFgCRCGACSCGEASCKEASCJEAaAJEIYAJIIYQBIIoQBIIkQBoAkQhgAkghhAEgihAEgiRAGgCRCGACSCGEASCKEASDJ0KwNM+MmT57caPnRo0c3LsOdd97ZaPlJkyY1LsPsYOONN268jrFjxzZaftttt21chvnmm6/xOmBOpCUMAEmEMAAkEcIAkEQIA0ASIQwASYQwACQRwgCQRAgDQBIhDABJhDAAJBHCAJBECANAEiEMAEmEMAAkEcIAkMTfEx6gm2++udHyxx57bOMy3HrrrY2Wf+yxxxqXYXYw//zzN17HmDFjGi0/bty4xmUYNmxY43UAObSEASCJEAaAJEIYAJIIYQBIIoQBIIkQBoAkQhgAkghhAEgihAEgiRAGgCRCGACSCGEASCKEASCJEAaAJEIYAJIIYQBIMjRrw+9U48ePT11+VjFy5MhGy48aNapxGYYMGdJo+YMPPrhxGUaMGNF4HcCcS0sYAJIIYQBIIoQBIIkQBoAkQhgAkghhAEgihAEgiRAGgCRCGACSCGEASCKEASCJEAaAJEIYAJIIYQBIIoQBIMmgsizLrI0DwJxMSxgAkghhAEgihAEgiRAGgCRCGACSCGEASCKEASCJEAaAJEIYAJIIYQBIIoQBIIkQBoAkQhgAkghhAEgihAEgiRAGgCRCGACSCGEASCKEASCJEAaAJEIYAJIIYQBIIoQBIIkQBoAkQhgAkghhAEgihAEgiRAGgCRCGACSCGEASCKEASCJEAaAJEIYAJIIYQBIIoQBIIkQBoAkQhgAkghhAEgihAEgiRAGgCRCGACSCGEASCKEASCJEAaAJEIYAJIIYQBIIoQBIIkQBoAkQhgAkghhAEgihAEgiRAGgCRCGACSCGEASCKEASCJEAaAJEIYAJIIYQBIIoQBoMjx/wBYhmNLul28twAAAABJRU5ErkJggg==", + "image/png": "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", "text/plain": [ - "
" + "
" ] }, "metadata": {}, @@ -225,7 +225,7 @@ "\n", "b. Analyse de la première instance:\n", " - Classe de la première instance: 5\n", - " - Type de la classe: \n", + " - Type de la classe: \n", " - Type de l'instance (données): \n", "\n", "=== Conclusion ===\n", @@ -253,7 +253,7 @@ "print(f\" ii. Image redimensionnée en 28x28: {image_2d.shape}\")\n", "\n", "# iii. Utiliser imshow avec cmap=mpl.cm.binary pour affichage en niveau de gris\n", - "plt.figure(figsize=(6, 6))\n", + "plt.figure()\n", "plt.imshow(image_2d, cmap=mpl.cm.binary)\n", "plt.title(\"Première instance du dataset MNIST\", fontsize=14)\n", "plt.axis('off') # Supprimer les axes pour une meilleure visualisation\n", @@ -337,9 +337,140 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 22, "id": "b0560d20", "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "=== Répartition des données MNIST ===\n", + "\n", + "Taille totale des données: 70000 échantillons\n", + "Taille totale des labels: 70000 échantillons\n", + "\n", + "=== Résultats de la répartition ===\n", + "Données d'apprentissage:\n", + " - X_train: (60000, 784) échantillons\n", + " - Y_train: (60000,) labels\n", + "\n", + "Données de test:\n", + " - X_test: (10000, 784) échantillons\n", + " - Y_test: (10000,) labels\n", + "\n", + "=== Vérification ===\n", + "Total apprentissage + test: 70000 échantillons\n", + "Cohérent avec le total original: True\n", + "\n", + "Répartition terminée!\n", + "60 000 échantillons pour l'apprentissage\n", + "10000 échantillons pour le test\n" + ] + } + ], + "source": [ + "print(\"=== Répartition des données MNIST ===\\n\")\n", + "\n", + "# Vérification de la taille totale des données\n", + "print(f\"Taille totale des données: {X.shape[0]} échantillons\")\n", + "print(f\"Taille totale des labels: {Y.shape[0]} échantillons\")\n", + "\n", + "# Répartition des données en une seule ligne (indexing sur ndarrays)\n", + "# a. Les 60 000 premières images composeront la base d'apprentissage\n", + "# b. Le reste des images constitue la base de test\n", + "X_train, X_test = X.iloc[:60000], X.iloc[60000:]\n", + "Y_train, Y_test = Y.iloc[:60000], Y.iloc[60000:]\n", + "\n", + "print(f\"\\n=== Résultats de la répartition ===\")\n", + "print(f\"Données d'apprentissage:\")\n", + "print(f\" - X_train: {X_train.shape} échantillons\")\n", + "print(f\" - Y_train: {Y_train.shape} labels\")\n", + "\n", + "print(f\"\\nDonnées de test:\")\n", + "print(f\" - X_test: {X_test.shape} échantillons\") \n", + "print(f\" - Y_test: {Y_test.shape} labels\")\n", + "\n", + "print(f\"\\n=== Vérification ===\")\n", + "print(f\"Total apprentissage + test: {X_train.shape[0] + X_test.shape[0]} échantillons\")\n", + "print(f\"Cohérent avec le total original: {X_train.shape[0] + X_test.shape[0] == X.shape[0]}\")\n", + "\n", + "print(f\"\\nRépartition terminée!\")\n", + "print(f\"60 000 échantillons pour l'apprentissage\")\n", + "print(f\"{X_test.shape[0]} échantillons pour le test\")" + ] + }, + { + "cell_type": "markdown", + "id": "6515e1aa", + "metadata": {}, + "source": [ + "# II- Apprentissage d'un classifieur binaire" + ] + }, + { + "cell_type": "markdown", + "id": "47d18544", + "metadata": {}, + "source": [ + "## 2 - Apprentissage des données" + ] + }, + { + "cell_type": "code", + "execution_count": 26, + "id": "bff45d57", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "=== Résultats de la répartition ===\n", + "\n", + "y_train_5: 0 True\n", + "1 False\n", + "2 False\n", + "3 False\n", + "4 False\n", + " ... \n", + "59995 False\n", + "59996 False\n", + "59997 True\n", + "59998 False\n", + "59999 False\n", + "Name: class, Length: 60000, dtype: bool\n", + "\n", + "y_test_5: 60000 False\n", + "60001 False\n", + "60002 False\n", + "60003 False\n", + "60004 False\n", + " ... \n", + "69995 False\n", + "69996 False\n", + "69997 False\n", + "69998 True\n", + "69999 False\n", + "Name: class, Length: 10000, dtype: bool\n" + ] + } + ], + "source": [ + "y_train_5 = (Y_train == 5)\n", + "y_test_5 = (Y_test == 5)\n", + "\n", + "print(f\"\\n=== Résultats de la répartition ===\")\n", + "print(f\"\\ny_train_5: {y_train_5}\")\n", + "print(f\"\\ny_test_5: {y_test_5}\")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "e02058de", + "metadata": {}, "outputs": [], "source": [] }