XOR_Example.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "f994e2b7",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline  \n",
    "#kod satırı Jupyter Notebook'a IDE'sine özel olarak yazılmış bir kod satırıdır. Bu kod satırı oluşturulan görselleri görmek için her defasında “plt.show()” satırını yazmayı önlüyor başka IDE'ler de çalışıyorsanız her defasında plt.show() yazmak gerekebilir."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "b4889e8c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(4, 2)\n"
     ]
    }
   ],
   "source": [
    "X = np.array([ [0, 1], [1, 0], [1, 1],[0, 0] ])\n",
    "y = np.array([ [1], [1], [0], [0]])\n",
    "print(np.shape(X))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "4a474866",
   "metadata": {},
   "outputs": [],
   "source": [
    "num_input = 2\n",
    "num_hidden = 5\n",
    "num_output = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "6e5c30ad",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 1.59920326  0.00344745 -1.25338253 -0.55145723 -0.0048261 ]\n",
      " [ 0.44043254 -1.50056345  1.0741207   0.52643556  0.42313882]]\n"
     ]
    }
   ],
   "source": [
    "Wxh = np.random.randn(num_input,num_hidden)\n",
    "print(Wxh)\n",
    "bh = np.zeros((1,num_hidden))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "5ca093eb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[-0.60236654]\n",
      " [ 0.16214038]\n",
      " [-1.33474827]\n",
      " [ 0.44096928]\n",
      " [ 1.13264757]]\n"
     ]
    }
   ],
   "source": [
    "Why = np.random.randn (num_hidden,num_output)\n",
    "print(Why)\n",
    "by = np.zeros((1,num_output))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "a422cceb",
   "metadata": {},
   "outputs": [],
   "source": [
    "def sigmoid(z):\n",
    "    return 1 / (1+np.exp(-z))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "994c8c86",
   "metadata": {},
   "outputs": [],
   "source": [
    "def sigmoid_derivative(z):\n",
    "    return np.exp(-z)/((1+np.exp(-z))**2)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "864c4080",
   "metadata": {},
   "outputs": [],
   "source": [
    "def forward_prop(X,Wxh,Why,bh,by):\n",
    "    z1 = np.dot(X,Wxh) + bh   # shape 4 x 5\n",
    "    a1 = sigmoid(z1)          # shape 4 x 5\n",
    "    z2 = np.dot(a1,Why) + by  # shape 4 x 1\n",
    "    y_hat = sigmoid(z2)       # shape 4 x 1\n",
    "    return z1,a1,z2,y_hat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "c4ee8cae",
   "metadata": {},
   "outputs": [],
   "source": [
    "def backword_prop(y,y_hat, z1, a1, z2,Why):\n",
    "    delta2 = np.multiply(-(y-y_hat),sigmoid_derivative(z2)) # shape 4 x 1\n",
    "    dJ_dWhy = np.dot(a1.T, delta2)                          # shape 5 x 1\n",
    "    delta1 = np.dot(delta2,Why.T)*sigmoid_derivative(z1)    # shape 4 x 5 \n",
    "    dJ_dWxh = np.dot(X.T, delta1)                           # shape 2 x 5\n",
    "    return dJ_dWxh, dJ_dWhy\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "dd3761db",
   "metadata": {},
   "outputs": [],
   "source": [
    "def cost_function(y, y_hat):\n",
    "    J = 0.5*sum((y-y_hat)**2)\n",
    "    return J"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "595288e5",
   "metadata": {},
   "outputs": [],
   "source": [
    "alpha = 0.01  # learning rate\n",
    "num_iterations = 50000\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "2bbf7b7d",
   "metadata": {},
   "outputs": [],
   "source": [
    "cost =[]\n",
    "for i in range(num_iterations):\n",
    "    z1,a1,z2,y_hat = forward_prop(X,Wxh,Why,bh,by) \n",
    "    dJ_dWxh, dJ_dWhy = backword_prop(y,y_hat, z1, a1, z2,Why)\n",
    "    #print(dJ_dWxh,dJ_dWhy)\n",
    "    #update weights\n",
    "    Wxh = Wxh - alpha * dJ_dWxh\n",
    "    Why = Why - alpha * dJ_dWhy\n",
    " \n",
    "    #print(y-y_hat)\n",
    "    #compute cost\n",
    "    c = cost_function(y, y_hat)\n",
    "    cost.append(c)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "16f5b77f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Cost')"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.grid()\n",
    "plt.plot(range(num_iterations),cost)\n",
    "plt.title('Cost Function')\n",
    "plt.xlabel('Training Iterations')\n",
    "plt.ylabel('Cost')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5ab68cfe",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dba309cc",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}