Perceptron_results_comparison.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "from Perceptron import Perceptron\n",
    "from LossFunction import LossFunction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Dans cet exemple, nous allons simuler le comportement de la porte logique OR sur 3 bits avec un entraînement 1000 itérations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# OR gate\n",
    "inputs_training = [\n",
    "    [0, 0, 0],  # 0\n",
    "    [0, 0, 1],  # 1\n",
    "    [0, 1, 0],  # 1\n",
    "    [0, 1, 1],  # 1\n",
    "    [1, 0, 0],  # 1\n",
    "    [1, 0, 1],  # 1\n",
    "    [1, 1, 0],  # 1\n",
    "    [1, 1, 1],  # 1\n",
    "]\n",
    "outputs_training = [0, 1, 1, 1, 1, 1, 1, 1]\n",
    "nb_iter = 10_000"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nous créons ensuite notre objet `Perceptron` avec les paramètres permettant d'entraîner le neurone. Également, nous utiliserons la fonction Tangente Hyperbolique pour l'activation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "p = Perceptron(inputs_training, outputs_training, nb_iter, LossFunction.TANH)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nous décidons ensuite d'afficher les informations pendant l'entraînement du neurone. Ainsi, nous aurons un visuel sur le pourcentage d'erreur du neurone à l'itération k. Nous afficherons ensuite l'évolution du taux d'erreur."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Pourcentage d'erreur (0) : 0.7851573815165486 -> 78.52%\n",
      "Pourcentage d'erreur (1000) : 0.005723198997329688 -> 0.57%\n",
      "Pourcentage d'erreur (2000) : 0.001439375670241531 -> 0.14%\n",
      "Pourcentage d'erreur (3000) : 0.0006037046811916397 -> 0.06%\n",
      "Pourcentage d'erreur (4000) : 0.0003233628302303177 -> 0.03%\n",
      "Pourcentage d'erreur (5000) : 0.00019924584914050088 -> 0.02%\n",
      "Pourcentage d'erreur (6000) : 0.0001343178192610241 -> 0.01%\n",
      "Pourcentage d'erreur (7000) : 9.636816424391431e-05 -> 0.01%\n",
      "Pourcentage d'erreur (8000) : 7.23657448548824e-05 -> 0.01%\n",
      "Pourcentage d'erreur (9000) : 5.626315272464317e-05 -> 0.01%\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8/fFQqAAAACXBIWXMAAAsTAAALEwEAmpwYAAAn8UlEQVR4nO3deZgdVZ3/8fenu7MRAgkQWbKQqEEMjCBGlnFjRtDAaHDcCG6gaEYl7o4GdfJDdOaHy+g4j6gERIQRIqLDRI3iqOCCignIYhICTViSsDUkQAiQpLu/80edm9S9qe6+3enqTvf9vJ7nPl3brfpW1e37vXXOqVOKCMzMrHE1DXYAZmY2uJwIzMwanBOBmVmDcyIwM2twTgRmZg3OicDMrME5EQwBkqZJCkktgx3LUCTpDEm/7+N7x0j6saTHJf2gv2PrZrsvk7R6gLZ1j6QTSt7GpyRdVNK6fybp9DLW3Z8kXSfp3YMdRxEnggKSnsy9OiU9nRt/624Q33WSjh/sOBrEG4H9gX0j4k1lbSQl+udWxiPidxHxvLK2100c50g6ZxfXcbykdflpEfFvEbHLX4Ipvv+qWfdJEfHdXV13wbYukfT5/l7v7si/MAtExJ6VYUn3AO+OiF8OXkT9Q1JLRLT3NG0Xt9EcER39tb7dwMHAHf15jMx2N74i6AVJR0v6o6THJD0g6euSRqZ5OxXf5C8FJX1T0g9z874g6VeSVLCdZklflvSIpDXAP3QTU5OkBZLukvSopCsl7VMT05mS7gN+nYpJrpf0VUmPAudIGpW2d5+khyR9S9KYtI6dilXyv17Tr6ZvSloqaTPwdwUxvlPSKkmbJK2R9E+5easkvSY33iKpTdJRafxYSX9Ix/yW/JWQpH0kfUfS/ZI2Srq6q+NUE8/XJK2V9ISkGyW9rIvlPgssBE5NV4NnpuP9GUn3SnpY0qWS9q453qenY/mIpE/n1tecikjuSsfiRklTJP02LXJL2s6ptb+qJT0/fZ4ek7RC0pzcvEsknS/pp2m9N0h6Tjf7//YU/6P5+AqWmyDpJ+l8bEzDk3Pzdzr+ksYCPwMO0o6r6IOU+yWvrChnfs22bpH0+u7Oj6TZwKdy5+OWND3/f9bn81MTzzzgrcAn0rZ+nKZX/tc2SVop6R9z7zlD0u+V/S9tlHS3pJNqVn2wsv+/TZJ+IWm/ro7/gIoIv7p5AfcAJ6ThFwHHkl1JTQNWAR9O86YBAbTk3nsd2dUEwB7AHcAZwMuAR4DJXWzzvcDtwBRgH+Da2nXnlv0Q8CdgMjAKuAC4oiamS4GxwJi0/XbgA2k/xgBfBZakbY0Dfgz8/7SOM4Df12wzgOem4UuAx4GXkP2wGF0Q4z8AzwEEvAJ4CjgqzVsIfK9m2VVpeBLwKHByWveJaXximv9T4PvABGAE8IoujmfVPgBvA/ZN+/8x4MGiuNOy5wD/lRt/F9AKPBvYE/gRcFnN8b4wHdcjgC3A89P8fwZuA56XjsURZEVOVcc0jR8PrEvDI9I2PwWMBP4e2AQ8L3cOHgWOTvv0PWBxF/szE3gSeHn6vHwlfR5OKFh2X+ANZJ/dccAPgKtz8wuPfz72ouMIvAO4viamx4BRPZ2f2vNR8H/W5/NTsP+XAJ+vmfYm4CCyz+OpwGbgwNznbBvwHqAZeB9wP6BcnHcBh6TtXwecN9jfcRHhRNDjAcolgoJ5Hwb+u+ZDVpgI0vgxwAbgXuC0brb5a+C9ufFX1a47N28V8Mrc+IHpw9iSi+nZuflnAPflxpU+zM/JTTsOuDu3fE+J4NJeHtOrgQ+l4eeSfantkca/ByxMw5+s/BPn3nsNcHraz05gQh3b22kfauZvBI7oYt45VCeCXwHvz40/r+B4T87N/zMwNw2vBk7pYjvdJYKXkX0ZNuXmXwGckzsHF+XmnQzc3sV2FpJLEmQ/ELbSxWe85r1HAhtzn7PC40/PiWBc+swdnMb/Fbi4nvNTez5q/8925fwUbPcSahJBwTI3V85p+py15ubtkbZ3QC7Oz+Tmvx/4eW/+d8p6uWioFyQdki6PH5T0BPBvQN2XdhFxA7CG7Mv3ym4WPQhYmxu/t5tlDwb+OxUZPEaWGDrIKjgr1ta8Jz8+kewDe2NuHT9P0+tVu/4qkk6S9CdJG9L6TyYdt4hoTTG/VtIewBzg8ty+vakSV3rvS8m+hKYAGyJiYy/irMTzcWVFUo+nde5N/efxIKrPx71kXzL54/1gbvgpsl+mpJjv6m28aZtrI6KzZruT6thm4boqIxGxmexqYieS9pB0QSpmeQL4LTBeUjO7cPwjYhPZ1cTcNOk0sh8Ale0O1vnpkaR3SLo593k8vCa27euOiKfS4J5F83u77TI5EfTON8mKbGZExF5kl+qVMv7N6e8eueUPyL9Z0llkl+P3A5/oZjsPkP2jVUztZtm1wEkRMT73Gh0R63PLRM178uOPAE8Dh+Xev3fsqDDfnN8nSVX71MX6t5M0Cvgh8GVg/4gYDyxlx3GD7NftacApwMqUHCr7dlnNvo2NiPPSvH0kje9q213E8zKyY/9msl+z48mKtnaqq+nC/WQJqmIqWdHKQ3W8dy1ZEVlv3Q9MkZT/f50KrO9i+e5UfbZS8t23i2U/RvaL+pj0eX955W10f/y7/DzkXAGcJuk4YDRZ8Wc956ende/K+alVtS1JB5MVK80nK9IbD/yV+j87uy0ngt4ZBzwBPCnpULIyQAAioo3sH/NtyioF30Xun17SIcDnyco/305WCXVkF9u5EvigpMmSJgALuonpW8C/pg8pkiZKOqXeHUq/Mi8EvirpWWkdkyS9Oi1yC3CYpCMljSa7NO+NkWTJrw1oT5Vnr6pZZnGa9j52XA0A/BfZlcKr0zEdrawSdXJEPEBWKfkNZZWaIyS9nJ6NI/tiaANaJC0E9urF/lwBfETSdEl7kl0Vfj/qa1V0EfA5STOUeYGkypfwQ2Tl2kVuIPv1+Im0n8cDryU7br11FfAaSS9V1tDhXLr+HhhH9iPhMWUNEP5fZUYPx/8hYN9KJW0XlpJ9YZ9LdvwqVzs9nZ+HgGk1STFvV85PrdpzMpYsObRB1giC7IpgyHMi6J2PA28hK9O+kKyiLO89ZBWCjwKHAX+ArCUM2ZfaFyLiloi4k+xq4rL0i7nWhWRl4bcAN5FVeHXla2QVvb+QtIms4viYXu7XJ8kq2P6UigB+SfZLkIi4g+yf9ZfAnUCvbsxKxQAfJEtuG8mO35KaZR4A/gj8LbljGhFrya4SPkX2z7eW7PhWPrdvJyv/vR14mKzOpifXkBV93UFWbPAMPRRt1bgYuIysmOTu9P4P1Pner5Adh1+Q/aD4NlmlIWQJ9rupyOHN+TdFxFayL/6TyK7gvgG8IyJu70XclXWtAM4iS7gPkJ2TdV0s/h8pvkfIPlc/r5lfePxTXFcAa9L+HFQQxxayz/UJVCf/ns5P5aa+RyXdVBDzrpyfWt8GZqZ9uDoiVgL/TvZZfQj4G+D6Pq57t1KpzTYzswblKwIzswbnRGBm1uCcCMzMGpwTgZlZgxtync7tt99+MW3atMEOw8xsSLnxxhsfiYjCG0WHXCKYNm0ay5cvH+wwzMyGFEld9lDgoiEzswZXaiKQNFvSakmtkna6O1bSVEnXSvqLpFslnVxmPGZmtrPSEkHqmOp8srshZ5L1KzKzZrHPAFdGxAvJOqD6RlnxmJlZsTKvCI4m65J1TbpFfjFZdwF5wY5+RPYm6zDKzMwGUJmVxZOo7iNkHTv3gXMOWR85HyDr0KnUB2ibmdnOBruy+DTgkoiYTNZH/WVFvQpKmidpuaTlbW1tAx6kmdlwVmYiWE91n/qT2bn/9DNJD2iJiD+S9Uu+0wMoImJRRMyKiFkTJ/bmeSlmZtaTMhPBMmBG6hd8JFll8JKaZe4DXgnZw7nJEkEpP/mX3bOBr/xiNVvbO3te2MysgZSWCNKDIOaT9S++iqx10ApJ50qakxb7GPAeSbeQ9V9+RpTUL/ZN927kP3/dSnunE4GZWV6pdxZHxFKyJxHlpy3MDa8EXlJmDGZm1r3BriwecH4Oj5lZtYZJBBryj5c2MytHwyQCMzMr1nCJwCVDZmbVGiYRCJcNmZkVaZhEYGZmxRouEZR0m4KZ2ZDVMInArYbMzIo1TCIwM7NiDZcIXDBkZlat4RKBmZlVcyIwM2twDZcI3GjIzKxawyQCudmQmVmhhkkEZmZWrPESgYuGzMyqNEwicMGQmVmxUhOBpNmSVktqlbSgYP5XJd2cXndIeqzMeMzMbGelPapSUjNwPnAisA5YJmlJejwlABHxkdzyHwBeWFY827fpsiEzsyplXhEcDbRGxJqI2AosBk7pZvnTyB5gXwo3GjIzK1ZmIpgErM2Nr0vTdiLpYGA68Osu5s+TtFzS8ra2tn4P1Myske0ulcVzgasioqNoZkQsiohZETFr4sSJAxyamdnwVmYiWA9MyY1PTtOKzKXEYqE831lsZlatzESwDJghabqkkWRf9ktqF5J0KDAB+GOJsbj5qJlZF0pLBBHRDswHrgFWAVdGxApJ50qak1t0LrA4/OgwM7NBUVrzUYCIWAosrZm2sGb8nDJj2CmmgdyYmdkQsLtUFpfOnc6ZmRVrmERgZmbFGi4RuCrCzKxawyQClwyZmRVrmERgZmbFGi4RuGDIzKxawyQClwyZmRVrmERgZmbFGi4RuNGQmVm1xkkEbjZkZlaocRKBmZkVarhE4EdVmplVa5hE4IIhM7NiDZMIzMysWOMlApcMmZlVaZhE4EZDZmbFSk0EkmZLWi2pVdKCLpZ5s6SVklZIurzMeMzMbGelPaFMUjNwPnAisA5YJmlJRKzMLTMDOBt4SURslPSssuKpcMmQmVm1Mq8IjgZaI2JNRGwFFgOn1CzzHuD8iNgIEBEPlxWM3G7IzKxQmYlgErA2N74uTcs7BDhE0vWS/iRpdtGKJM2TtFzS8ra2tpLCNTNrTN0mAklNkt5c4vZbgBnA8cBpwIWSxtcuFBGLImJWRMyaOHHiLm3QfQ2ZmVXrNhFERCfwiT6uez0wJTc+OU3LWwcsiYhtEXE3cAdZYuh3bjVkZlasnqKhX0r6uKQpkvapvOp43zJghqTpkkYCc4ElNctcTXY1gKT9yIqK1tQdvZmZ7bJ6Wg2dmv6elZsWwLO7e1NEtEuaD1wDNAMXR8QKSecCyyNiSZr3KkkrgQ7gnyPi0d7uRG+4ryEzs2o9JoKImN7XlUfEUmBpzbSFueEAPppepXLJkJlZsR4TgaR3FE2PiEv7PxwzMxto9RQNvTg3PBp4JXATMCQTgVsNmZlVq6do6AP58dS8c3FZAZXFrYbMzIr15YayzUCf6w3MzGz3Uk8dwY/Z0UVPEzATuLLMoMrkkiEzs2r11BF8OTfcDtwbEetKiqc07mvIzKxYj0VDEfEb4B5gRERcDzwqaVzZgZmZ2cDoMRFIeg9wFXBBmjSZ7I7gISncbMjMrEo9lcVnAS8BngCIiDuB0p8b0O9cMmRmVqieRLAlPU8AAEktuM7VzGzYqCcR/EbSp4Axkk4EfgD8uNywyuOSITOzavUkgk8CbcBtwD+R9R30mTKDKoNLhszMinXbfDQ9d3hFRBwKXDgwIZmZ2UDq6cE0HcBqSVMHKB4zMxtg9dxQNgFYIenPZN1LABARc0qLyszMBkw9ieBfSo9iAMi9zpmZFaqnjuCCVEfQa5JmA18je0LZRRFxXs38M4AvseNZxl+PiIv6si0zM+ubbhNBRHRIWi1pakTc15sVpyRyPnAi2UPql0laEhEraxb9fkTM71XUu8DNR83MqpVZR3A00BoRawAkLQZOAWoTwYBwwZCZWbEy6wgmAWtz4+uAYwqWe4OklwN3AB+JiLW1C0iaB8wDmDrVDZjMzPpTb3sf/Q2wjOxRlf3hx8C0iHgB8L/Ad7uIYVFEzIqIWRMnTtylDYZ7xzAzq9KX3kcnUV/vo+uBKbnxyeyoFAYgIh6NiC1p9CLgRXWst0/caMjMrFiZvY8uA2ZImi5pJDAXWJJfQNKBudE5wKp6gjYzs/5TTx3BlojYWmmHX2/voxHRLmk+cA1Z89GLI2KFpHOB5RGxBPigpDlkTz7bAJzRt92on1sNmZlVqycR1PY++n7q7H00IpaSdVKXn7YwN3w2cHb94fadi4bMzIrVUzS0gGHQ+6iZmRXr8YogIjrJeh4dFr2PumTIzKxal4lA0rVk35sbIuKNAxdSOeRbyszMCnV3RXBG+tsxAHGYmdkg6TIRRMS9AxnIQAk3GzIzq9Jd0dAmuilSj4i9SomoJG41ZGZWrLsrgnEAkj4HPABcRtZ321uBA7t6n5mZDS31NB+dExHfiIhNEfFERHyTrBfRIckFQ2Zm1epJBJslvVVSs6QmSW8l1x21mZkNbfUkgrcAbwYeSq83pWlmZjYM1HND2T0M4aKgWm40ZGZWrZ4rgmHBD683MyvWMInAzMyKNWAicNmQmVlePU8o21/StyX9LI3PlHRm+aH1LxcMmZkVq+eK4BKyh8sclMbvAD5cUjxmZjbA6kkE+0XElUAnZE8eo86O6CTNlrRaUqukBd0s9wZJIWlWXVHvArcaMjOrVu8NZfuSCtclHQs83tObJDUD5wMnATOB0yTNLFhuHPAh4IZexN1rbjRkZlasnkTwUbKHzj9H0vXApcAH6njf0UBrRKyJiK3AYorvR/gc8AXgmfpCNjOz/tRjIoiIm4BXAH9L9qjKwyLi1jrWPQlYmxtfl6ZtJ+koYEpE/LS7FUmaJ2m5pOVtbW11bLprLhkyM6vW453Fkl5fM+kQSY8Dt0XEw33dsKQm4CvseABOlyJiEbAIYNasWX36LvcTyszMivWYCIAzgeOAa9P48cCNwHRJ50bEZV28bz0wJTc+OU2rGAccDlyX7vo9AFgiaU5ELK97D8zMbJfUkwhagOdHxEOQ3VdAVk9wDPBbsucUFFkGzJA0nSwBzCXXWV1EPA7sVxmXdB3w8bKTgFsNmZlVq6eyeEolCSQPp2kbgG1dvSk1M51Pdg/CKuDKiFgh6VxJc3Yl6L5wqyEzs2L1XBFcJ+knwA/S+BvStLHAY929MSKWAktrpi3sYtnj64jFzMz6WT2J4CyyL/+XpPFLgR9G9hT4vysrsLKE2w2ZmVWp53kEAVyVXkOWS4bMzIrV0+ncsZKWSXpS0lZJHZKeGIjgzMysfPVUFn8dOA24ExgDvJus64ghya2GzMyq1fU8gohoBZojoiMivgPMLjes/udWQ2ZmxeqpLH5K0kjgZklfBB6gIR9oY2Y2PNXzhf72tNx8YDPZ3cK13U6YmdkQVU8ieF1EPBMRT0TEZyPio8Bryg6sLK4jMDOrVk8iOL1g2hn9HMcAcCWBmVmRLusIJJ1G1jfQdElLcrPGARvKDszMzAZGd5XFfyCrGN4P+Pfc9E1APc8j2C35zmIzs2pdJoKIuBe4l6wL6iHPzUfNzIrVc2fx6yXdKelxSU9I2uQ7i83Mho967iP4IvDaiFhVdjADwa2GzMyq1dNq6KHhkARcMmRmVqyeK4Llkr4PXA1sqUyMiB+VFZSZmQ2ceq4I9gKeAl4FvDa96rqhTNJsSasltUpaUDD/vZJuk3SzpN9Lmtmb4M3MbNfV8zyCd/ZlxZKayXopPRFYByyTtCQiVuYWuzwivpWWnwN8hZI6tJObDZmZFaqn1dAhkn4l6a9p/AWSPlPHuo8GWiNiTURsBRYDp+QXiIh866Ox4Eb+ZmYDrZ6ioQuBs0kPqo+IW4G5dbxvErA2N74uTasi6SxJd5G1Tvpg0YokzZO0XNLytra2OjbdNbcaMjOrVk8i2CMi/lwzrb2/AoiI8yPiOcAngcIrjYhYFBGzImLWxIkT+7QdFwyZmRWrJxE8Iuk5pGIbSW8k63qiJ+vJuqyumJymdWUx8Lo61mtmZv2onuajZwGLgEMlrQfuBt5Wx/uWATMkTSdLAHPJOrHbTtKMiLgzjf4D2eMwS+W+hszMqtXTamgNcIKksUBTRGyqZ8UR0S5pPnAN0AxcHBErJJ0LLI+IJcB8SSeQ1T9spLjL637hRkNmZsV6TASS/g34YkQ8lsYnAB+LiB5bDkXEUmBpzbSFueEP9TZgMzPrX/XUEZxUSQIAEbEROLm0iErmVkNmZtXqSQTNkkZVRiSNAUZ1s/xuyUVDZmbF6qks/h7wK0nfSePvBL5bXkhmZjaQuk0EyvpluBy4BTghTf5cRFxTdmBlccmQmVm1bhNBRISkpRHxN8DPByimUsi3lJmZFaqnjuAmSS8uPRIzMxsU9dQRHAO8VdK9wGay3hoiIl5QamQlCTcbMjOrUk8ieHXpUQwElwyZmRWqJxH4J7SZ2TBWTyL4KVkyEDAamA6sBg4rMa7SOKuZmVWrp6+hv8mPSzoKeH9pEZXEJUNmZsXqaTVUJSJuIqtANjOzYaCeTuc+mhttAo4C7i8topK50ZCZWbV66gjG5YbbyeoMflhOOOXxw+vNzIrVU0fwWQBJe6bxJ8sOyszMBk6PdQSSDpf0F2AFsELSjZIOLz+0srhsyMwsr57K4kXARyPi4Ig4GPhYmtYjSbMlrZbUKmlBwfyPSlop6VZJv5J0cO/Cr19TKhlyHYGZWbV6EsHYiLi2MhIR1wFje3qTpGbgfOAkYCZwmqSZNYv9BZiVuqu4CvhinXH3WlOqI+h0IjAzq1JPIlgj6V8kTUuvzwBr6njf0UBrRKyJiK3AYuCU/AIRcW1EPJVG/wRM7k3wvVGpK+70JYGZWZV6EsG7gInAj8haC+2XpvVkErA2N74uTevKmcDP6lhvn+y4InAiMDPL67LVkKTRwHuB5wK3kT2wflsZQUh6GzALeEUX8+cB8wCmTp3ap21UEoHzgJlZte6uCL5L9uV8G1k5/5d6ue71wJTc+OQ0rYqkE4BPA3MiYkvRiiJiUUTMiohZEydO7GUYmSYXDZmZFeruPoKZlX6GJH0b+HMv170MmCFpOlkCmAu8Jb+ApBcCFwCzI+LhXq6/V3bUEZS5FTOzoae7K4LtxUAR0d7bFaf3zAeuAVYBV0bECknnSpqTFvsSsCfwA0k3S1rS2+3US9uLhpwJzMzyursiOELSE2lYwJg0XnlC2V49rTwilgJLa6YtzA2f0PuQ+8Z1BGZmxbpMBBHRPJCBlM11BGZmxXrdDfVQ5RvKzMyKNUwi8A1lZmbFGiYRNLmy2MysUMMlAhcNmZlVa6BEkP110ZCZWbWGSQTyFYGZWaGGSQQ7nkfgTGBmltdAicC9j5qZFWm8RNA5yIGYme1mGiYR+D4CM7NiDZcInAfMzKo1TCJwHYGZWbGGSwROA2Zm1RooEWR/fUVgZlatYRKBbygzMyvWMInAN5SZmRUrNRFImi1ptaRWSQsK5r9c0k2S2iW9scxYdtxH4ERgZpZXWiKQ1AycD5wEzAROkzSzZrH7gDOAy8uKo6KSCDqcB8zMqnT3zOJddTTQGhFrACQtBk4BVlYWiIh70rzS7/cd2ZLlvG0dvrXYzCyvzKKhScDa3Pi6NK3XJM2TtFzS8ra2tj4FMyolgme2dfTp/WZmw9WQqCyOiEURMSsiZk2cOLFP62hqEiObm9jS7isCM7O8MhPBemBKbnxymjZoRrU0sWWbE4GZWV6ZiWAZMEPSdEkjgbnAkhK316NRI5p4pt1FQ2ZmeaUlgohoB+YD1wCrgCsjYoWkcyXNAZD0YknrgDcBF0haUVY8AKNamn1FYGZWo8xWQ0TEUmBpzbSFueFlZEVGA8JXBGZmOxsSlcX9ZdzoETzx9LbBDsPMbLfSUIngwL1G88Djzwx2GGZmu5WGSgQHjR/D+o1Ps9VNSM3MtmuoRPC3z9mXp7d1cPkN9w52KGZmu42GSgR/d+izOP55Eznnxyu59I/3DHY4Zma7hYZKBM1N4ltvexEnPH9/Fv7PCs6/ttXdUptZw2uoRAAwekQz33zbUbzuyIP40jWrOe/ntzsZmFlDK/U+gt3ViOYmvvLmIxk7qoULfrOGTc+087lTDqe58vQaM7MG0pCJALJO6D7/usMZN3oE3/rNXWze0s6X33QEI5ob7iLJzBpcwyYCyJ5jvOCkQ9lrTAtf/PlqNm9p5+tvOYrRI5oHOzQzswHjn7/A+49/Lp875TB+ueph3vmdZWx6xncfm1njcCJI3n7cNL566hH8+Z4NzP6P3/GH1kcGOyQzswHhRJDzjy+czPfnHcvIlibectENnHX5Taxpe3KwwzIzK5WGWtPJWbNmxfLly0vdxtNbO/jGda1c9Lu72drRyezDDuDtxx3MMdP3QXLLIjMbeiTdGBGzCuc5EXStbdMWFv32Lq5cvo7Hn97G5AljmH3YAZw4c3+OnDqeUS2uVDazocGJYBc9vbWDn9x6P0tve4DrWx9la0cno1qaOGLKeGYdPIFDD9yL5+0/jun7jWVki0vbzGz3M2iJQNJs4GtAM3BRRJxXM38UcCnwIuBR4NSIuKe7dQ5GIsjb9Mw2rm99lGX3bGDZPRtYcf8TdHRmx7ClSUyeMIaDxo9h0vjs70HjR7PP2FFM2GMEE8aOZJ89RrL3mBE0+eY1MxtA3SWC0u4jkNQMnA+cCKwDlklaEhErc4udCWyMiOdKmgt8ATi1rJj6w7jRI5h9+AHMPvwAALa0d3D3I5tZ/eAmVj+4ifs2PMX9jz3Nb+9s4+FNWyjKs02CPUe1sOeoFvYY1cLYkc3sMbKFsaNaGDsqGx7V0sTIliZGNmd/R6S/2TSlv82MaBYtzaJJorkpvdJwU354+3xobmqiWaKpie3LNzUJkd1bIaBJAkH6s326lOZRmafCZVyXYjZ0lHlD2dFAa0SsAZC0GDgFyCeCU4Bz0vBVwNclKYZQedWolmYOPWAvDj1gr53mbW3v5OFNz7Bx8zY2PLWVjZu3siG9Nj2zjc1bO3hqazubt2R/1z/29PbxLe0dbOvoZGt7J51D5mhUK0wQlcTRRRKhJn+oan3VM2tzTW3q2Wn5gvi6mtvzumvnd/3+nuKqtdO6u9lWT3Huqv5O6KX8POjnle7Ox/BDr5zBa484qN/WV1FmIpgErM2NrwOO6WqZiGiX9DiwL1DViF/SPGAewNSpU8uKt9+NbGli8oQ9mDxh19bT3tHJto5ga3snWzvSq71ze6Jo7ww6OoPOCNo7sr8dnUFHBJ25eR2d0BFBR2cnHZ1k82LH/AiICAKIgM6UjyMgSPN3mlc9fftyufXUvrdqmZrped39HKj9rVC7aO17u1t3T++tXWKndXezrZ3n9S7O7kZ7Oga7qr9/jpXxe6a/fzP2e4z9vMK9x4zo3xUmQ6KLiYhYBCyCrI5gkMMZcC3NTbQ0w5iRbqVkZv2vzCYu64EpufHJaVrhMpJagL3JKo3NzGyAlJkIlgEzJE2XNBKYCyypWWYJcHoafiPw66FUP2BmNhyUVjSUyvznA9eQNR+9OCJWSDoXWB4RS4BvA5dJagU2kCULMzMbQKXWEUTEUmBpzbSFueFngDeVGYOZmXXPt8GamTU4JwIzswbnRGBm1uCcCMzMGtyQ631UUhtwbx/fvh81dy03AO9zY/A+N4Zd2eeDI2Ji0Ywhlwh2haTlXfW+N1x5nxuD97kxlLXPLhoyM2twTgRmZg2u0RLBosEOYBB4nxuD97kxlLLPDVVHYGZmO2u0KwIzM6vhRGBm1uAaJhFImi1ptaRWSQsGO56+kjRF0rWSVkpaIelDafo+kv5X0p3p74Q0XZL+M+33rZKOyq3r9LT8nZJO72qbuwtJzZL+IuknaXy6pBvSvn0/dXeOpFFpvDXNn5Zbx9lp+mpJrx6kXamLpPGSrpJ0u6RVko4b7udZ0kfS5/qvkq6QNHq4nWdJF0t6WNJfc9P67bxKepGk29J7/lOq41mZ2eMCh/eLrBvsu4BnAyOBW4CZgx1XH/flQOCoNDwOuAOYCXwRWJCmLwC+kIZPBn5G9ijWY4Eb0vR9gDXp74Q0PGGw96+Hff8ocDnwkzR+JTA3DX8LeF8afj/wrTQ8F/h+Gp6Zzv0oYHr6TDQP9n51s7/fBd6dhkcC44fzeSZ7dO3dwJjc+T1juJ1n4OXAUcBfc9P67bwCf07LKr33pB5jGuyDMkAH/jjgmtz42cDZgx1XP+3b/wAnAquBA9O0A4HVafgC4LTc8qvT/NOAC3LTq5bb3V5kT7j7FfD3wE/Sh/wRoKX2HJM9A+O4NNySllPtec8vt7u9yJ7WdzepQUft+RuO55kdzzDfJ523nwCvHo7nGZhWkwj65bymebfnplct19WrUYqGKh+winVp2pCWLoVfCNwA7B8RD6RZDwL7p+Gu9n2oHZP/AD4BdKbxfYHHIqI9jefj375vaf7jafmhtM/TgTbgO6k47CJJYxnG5zki1gNfBu4DHiA7bzcyvM9zRX+d10lpuHZ6txolEQw7kvYEfgh8OCKeyM+L7KfAsGkXLOk1wMMRceNgxzKAWsiKD74ZES8ENpMVGWw3DM/zBOAUsiR4EDAWmD2oQQ2CwTivjZII1gNTcuOT07QhSdIIsiTwvYj4UZr8kKQD0/wDgYfT9K72fSgdk5cAcyTdAywmKx76GjBeUuUpe/n4t+9bmr838ChDa5/XAesi4oY0fhVZYhjO5/kE4O6IaIuIbcCPyM79cD7PFf11Xten4drp3WqURLAMmJFaH4wkq1haMsgx9UlqAfBtYFVEfCU3awlQaTlwOlndQWX6O1Lrg2OBx9Ml6DXAqyRNSL/EXpWm7XYi4uyImBwR08jO3a8j4q3AtcAb02K1+1w5Fm9My0eaPje1NpkOzCCrWNvtRMSDwFpJz0uTXgmsZBifZ7IioWMl7ZE+55V9HrbnOadfzmua94SkY9MxfEduXV0b7EqTAaycOZmshc1dwKcHO55d2I+Xkl023grcnF4nk5WN/gq4E/glsE9aXsD5ab9vA2bl1vUuoDW93jnY+1bn/h/PjlZDzyb7B28FfgCMStNHp/HWNP/Zufd/Oh2L1dTRmmKQ9/VIYHk611eTtQ4Z1ucZ+CxwO/BX4DKylj/D6jwDV5DVgWwju/I7sz/PKzArHb+7gK9T0+Cg6OUuJszMGlyjFA2ZmVkXnAjMzBqcE4GZWYNzIjAza3BOBGbdkPQWSVMHOw6zMjkRWMOS9GT6O03SWwrmnwk8KyLu68O6P1Uz/oc+B2pWMjcftYYl6cmI2FPS8cDHI+I1vXhvS+zo/6bLdfdDmGal8xWBGZwHvEzSzak//GZJX5K0LPUB/08Ako6X9DtJS8jueEXS1ZJuVNaH/rw07TxgTFrf99K0ytWH0rr/mvqMPzW37uu04/kD36urH3mzftDS8yJmw94CclcE6Qv98Yh4saRRwPWSfpGWPQo4PCLuTuPviogNksYAyyT9MCIWSJofEUcWbOv1ZHcMHwHsl97z2zTvhcBhwP3A9WT97Py+v3fWrJavCMx29iqy/l1uJuvie1+y/moA/pxLAgAflHQL8CeyTsBm0L2XAldEREdEPAT8Bnhxbt3rIqKTrOuQaf2wL2Y98hWB2c4EfCAiqjpnS3UJm2vGTyB76MlTkq4j6/+mr7bkhjvw/6cNEF8RmMEmssd+VlwDvC91942kQ9JDYWrtDWxMSeBQsscDVmyrvL/G74BTUz3ERLLHFu7uPWPaMOdfHGZZ754dqYjnErJnHUwDbkoVtm3A6wre93PgvZJWkfVy+afcvEXArZJuiqzL7Ir/Jnvc4i1kvch+IiIeTInEbFC4+aiZWYNz0ZCZWYNzIjAza3BOBGZmDc6JwMyswTkRmJk1OCcCM7MG50RgZtbg/g/LVVF8UVU7KgAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "p.train(display_loss=True)\n",
    "p.display_mean_squared_loss_graph()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Le bloc ci-dessous permet d'effectuer une prédiction avec notre neurone entraîné sur l'entrée [0, 0, 1]. En tant qu'humain, nous savons en appliquant la porte logique OR sur les bits 0, 0 et 1, le résultat sera 1.\n",
    "\n",
    "Est-ce que notre neurone fraîchement entraîné saura trouver la bonne valeur ?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Output: 0.9945843559585799\n"
     ]
    }
   ],
   "source": [
    "prob = p.guess([0, 0, 1])\n",
    "print(f\"Output: {prob}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Comme nous pouvons le voir dans le résultat du bloc ci-dessus, le résultat retourné par notre neurone est 0.945071... ce qui signifie que notre Perceptron a une probabilité de 94.5% de s'activer. Or, la probabilité augmente lorsque notre neurone \"pense\" que le résultat est 1. A l'inverse, la probabilité diminue lorsque notre neurone pense que la réponse est 0.\n",
    "\n",
    "Nous remarquons donc que la probabilité est plutôt élevée ce qui signifie que notre neurone pense que le résultat de la porte logique OR sur les bits 0, 0 et 1 est 1.\n",
    "\n",
    "Essayons maintenant sur les bits 0, 0 et 0. Si tout se passe bien, notre neurone devrait retourner une probabilité faible."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Output: 0.016478701638666323\n"
     ]
    }
   ],
   "source": [
    "prob = p.guess([0, 0, 0])\n",
    "print(f\"Output: {prob}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Effectivement, notre neurone retourne la valeur 0.1913... ce qui est bien plus faible comparé à la probabilité précédente.\n",
    "\n",
    "On en déduit donc que notre neurone a été entraîné convenablement pour ce problème donné. Cet entraînement a été rendu possible, car les données de la porte logique OR sur 3 bits sont linéairement séparable. Autrement-dit, nous pouvons placer un hyperplan de manière à séparer les entrées pour lesquelles le résultat est 0 et celles pour lesquelles le résultat est 1."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Essayons maintenant avec la porte logique XOR sur 2 bits avec le même nombre d'itérations pour l'entraînement, mais cette fois-ci avec la fonction Sigmoïde pour l'activation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Pourcentage d'erreur (0) : 0.25125467684279656 -> 25.13%\n",
      "Pourcentage d'erreur (1000) : 0.2512545981885783 -> 25.13%\n",
      "Pourcentage d'erreur (2000) : 0.2512546741278102 -> 25.13%\n",
      "Pourcentage d'erreur (3000) : 0.2512546949996377 -> 25.13%\n",
      "Pourcentage d'erreur (4000) : 0.2512546993982512 -> 25.13%\n",
      "Pourcentage d'erreur (5000) : 0.25125470030218683 -> 25.13%\n",
      "Pourcentage d'erreur (6000) : 0.25125470048843435 -> 25.13%\n",
      "Pourcentage d'erreur (7000) : 0.25125470052694887 -> 25.13%\n",
      "Pourcentage d'erreur (8000) : 0.25125470053492915 -> 25.13%\n",
      "Pourcentage d'erreur (9000) : 0.25125470053658416 -> 25.13%\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# XOR gate\n",
    "inputs_training = [\n",
    "    [0, 0],  # 0\n",
    "    [0, 1],  # 1\n",
    "    [1, 0],  # 1\n",
    "    [1, 1],  # 0\n",
    "]\n",
    "outputs_training = [0, 1, 1, 0]\n",
    "\n",
    "p = Perceptron(inputs_training, outputs_training, nb_iter, LossFunction.SIGMOID)\n",
    "\n",
    "p.train(display_loss=True)\n",
    "p.display_mean_squared_loss_graph()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nous remarquons assez facilement que le taux d'erreur est bien différent du précédent. Ceci est dû au fait que la porte logique XOR n'est pas linéairement séparable. De ce fait, un seul neurone n'est pas suffisant pour résoudre ce problème."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Output: 0.4999999999568976\n"
     ]
    }
   ],
   "source": [
    "prob = p.guess([0, 1])\n",
    "print(f\"Output: {prob}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Comme nous pouvons le voir avec le résultat retourné par notre neurone sur l'entrée 0 et 1, celui-ci est bien plus mitigé contrairement aux résultats précédents."
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "2893b4f6329644dcc5fbc6c44db0bdd3116d568eb12d76f7a2ce41daa65cea55"
  },
  "kernelspec": {
   "display_name": "Python 3.9.7 64-bit ('venv': venv)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}