NN _from_scratch.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "0d00069e-72a4-4b0a-aa7f-28a15dc73c66",
   "metadata": {},
   "source": [
    "# Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "5fbc5974-0b66-4c73-8d45-d4cd75c32c61",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from matplotlib import pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "58967622-76ac-489b-9156-543f50b0a6e3",
   "metadata": {},
   "source": [
    "# Load Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "b4382d32-35b3-4f42-9cfa-04660dde1b6c",
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.read_csv('digit_recognizer.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "208a9816-54bf-42d9-81c5-240a856e5a6e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>label</th>\n",
       "      <th>pixel0</th>\n",
       "      <th>pixel1</th>\n",
       "      <th>pixel2</th>\n",
       "      <th>pixel3</th>\n",
       "      <th>pixel4</th>\n",
       "      <th>pixel5</th>\n",
       "      <th>pixel6</th>\n",
       "      <th>pixel7</th>\n",
       "      <th>pixel8</th>\n",
       "      <th>...</th>\n",
       "      <th>pixel774</th>\n",
       "      <th>pixel775</th>\n",
       "      <th>pixel776</th>\n",
       "      <th>pixel777</th>\n",
       "      <th>pixel778</th>\n",
       "      <th>pixel779</th>\n",
       "      <th>pixel780</th>\n",
       "      <th>pixel781</th>\n",
       "      <th>pixel782</th>\n",
       "      <th>pixel783</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 785 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   label  pixel0  pixel1  pixel2  pixel3  pixel4  pixel5  pixel6  pixel7  \\\n",
       "0      1       0       0       0       0       0       0       0       0   \n",
       "1      0       0       0       0       0       0       0       0       0   \n",
       "2      1       0       0       0       0       0       0       0       0   \n",
       "3      4       0       0       0       0       0       0       0       0   \n",
       "4      0       0       0       0       0       0       0       0       0   \n",
       "\n",
       "   pixel8  ...  pixel774  pixel775  pixel776  pixel777  pixel778  pixel779  \\\n",
       "0       0  ...         0         0         0         0         0         0   \n",
       "1       0  ...         0         0         0         0         0         0   \n",
       "2       0  ...         0         0         0         0         0         0   \n",
       "3       0  ...         0         0         0         0         0         0   \n",
       "4       0  ...         0         0         0         0         0         0   \n",
       "\n",
       "   pixel780  pixel781  pixel782  pixel783  \n",
       "0         0         0         0         0  \n",
       "1         0         0         0         0  \n",
       "2         0         0         0         0  \n",
       "3         0         0         0         0  \n",
       "4         0         0         0         0  \n",
       "\n",
       "[5 rows x 785 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "aff6b92c-c61f-4cd8-a3cf-bec919a554dd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 42000 entries, 0 to 41999\n",
      "Columns: 785 entries, label to pixel783\n",
      "dtypes: int64(785)\n",
      "memory usage: 251.5 MB\n"
     ]
    }
   ],
   "source": [
    "data.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1b5f7b0b-1126-4c7a-a670-a7cf8c9229ec",
   "metadata": {},
   "source": [
    "# Split Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "f17151a9-7052-40ac-8a41-53ab513ab8d4",
   "metadata": {},
   "outputs": [],
   "source": [
    "data = np.array(data)\n",
    "m, n = data.shape\n",
    "np.random.shuffle(data)\n",
    "\n",
    "data_dev = data[0:1000].T\n",
    "Y_dev = data_dev[0]\n",
    "X_dev = data_dev[1:n]\n",
    "X_dev = X_dev / 255.\n",
    "\n",
    "data_train = data[1000:m].T\n",
    "Y_train = data_train[0]\n",
    "X_train = data_train[1:n]\n",
    "X_train = X_train / 255.\n",
    "_,m_train = X_train.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3325f479-e46b-4580-94e6-4c56afbe89e3",
   "metadata": {},
   "source": [
    "# Mathematical Intution\n",
    "\n",
    "\n",
    "\n",
    "The Neural Network will have a simple two-layer architecture. Input layer $a^{[0]}$ will have 784 units corresponding to the 784 pixels in each 28x28 input image. A hidden layer $a^{[1]}$ will have 10 units with ReLU activation, and finally our output layer $a^{[2]}$ will have 10 units corresponding to the ten digit classes with softmax activation.\n",
    "\n",
    "**Forward propagation**\n",
    "\n",
    "$$Z^{[1]} = W^{[1]} X + b^{[1]}$$\n",
    "$$A^{[1]} = g_{\\text{ReLU}}(Z^{[1]}))$$\n",
    "$$Z^{[2]} = W^{[2]} A^{[1]} + b^{[2]}$$\n",
    "$$A^{[2]} = g_{\\text{softmax}}(Z^{[2]})$$\n",
    "\n",
    "**Backward propagation**\n",
    "\n",
    "$$dZ^{[2]} = A^{[2]} - Y$$\n",
    "$$dW^{[2]} = \\frac{1}{m} dZ^{[2]} A^{[1]T}$$\n",
    "$$dB^{[2]} = \\frac{1}{m} \\Sigma {dZ^{[2]}}$$\n",
    "$$dZ^{[1]} = W^{[2]T} dZ^{[2]} .* g^{[1]\\prime} (z^{[1]})$$\n",
    "$$dW^{[1]} = \\frac{1}{m} dZ^{[1]} A^{[0]T}$$\n",
    "$$dB^{[1]} = \\frac{1}{m} \\Sigma {dZ^{[1]}}$$\n",
    "\n",
    "**Parameter updates**\n",
    "\n",
    "$$W^{[2]} := W^{[2]} - \\alpha dW^{[2]}$$\n",
    "$$b^{[2]} := b^{[2]} - \\alpha db^{[2]}$$\n",
    "$$W^{[1]} := W^{[1]} - \\alpha dW^{[1]}$$\n",
    "$$b^{[1]} := b^{[1]} - \\alpha db^{[1]}$$\n",
    "\n",
    "**Vars and shapes**\n",
    "\n",
    "Forward prop\n",
    "\n",
    "- $A^{[0]} = X$: 784 x m\n",
    "- $Z^{[1]} \\sim A^{[1]}$: 10 x m\n",
    "- $W^{[1]}$: 10 x 784 (as $W^{[1]} A^{[0]} \\sim Z^{[1]}$)\n",
    "- $B^{[1]}$: 10 x 1\n",
    "- $Z^{[2]} \\sim A^{[2]}$: 10 x m\n",
    "- $W^{[1]}$: 10 x 10 (as $W^{[2]} A^{[1]} \\sim Z^{[2]}$)\n",
    "- $B^{[2]}$: 10 x 1\n",
    "\n",
    "Backprop\n",
    "\n",
    "- $dZ^{[2]}$: 10 x m ($~A^{[2]}$)\n",
    "- $dW^{[2]}$: 10 x 10\n",
    "- $dB^{[2]}$: 10 x 1\n",
    "- $dZ^{[1]}$: 10 x m ($~A^{[1]}$)\n",
    "- $dW^{[1]}$: 10 x 10\n",
    "- $dB^{[1]}$: 10 x 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "09d98171-c000-4165-bad9-e20fa0a5083c",
   "metadata": {},
   "outputs": [],
   "source": [
    "def init_params():\n",
    "    W1 = np.random.rand(10, 784) - 0.5\n",
    "    b1 = np.random.rand(10, 1) - 0.5\n",
    "    W2 = np.random.rand(10, 10) - 0.5\n",
    "    b2 = np.random.rand(10, 1) - 0.5\n",
    "    return W1, b1, W2, b2\n",
    "\n",
    "def ReLU(Z):\n",
    "    return np.maximum(Z, 0)\n",
    "\n",
    "def softmax(Z):\n",
    "    A = np.exp(Z) / sum(np.exp(Z))\n",
    "    return A\n",
    "    \n",
    "def forward_prop(W1, b1, W2, b2, X):\n",
    "    Z1 = W1.dot(X) + b1\n",
    "    A1 = ReLU(Z1)\n",
    "    Z2 = W2.dot(A1) + b2\n",
    "    A2 = softmax(Z2)\n",
    "    return Z1, A1, Z2, A2\n",
    "\n",
    "def ReLU_deriv(Z):\n",
    "    return Z > 0\n",
    "\n",
    "def one_hot(Y):\n",
    "    one_hot_Y = np.zeros((Y.size, Y.max() + 1))\n",
    "    one_hot_Y[np.arange(Y.size), Y] = 1\n",
    "    one_hot_Y = one_hot_Y.T\n",
    "    return one_hot_Y\n",
    "\n",
    "def backward_prop(Z1, A1, Z2, A2, W1, W2, X, Y):\n",
    "    one_hot_Y = one_hot(Y)\n",
    "    dZ2 = A2 - one_hot_Y\n",
    "    dW2 = 1 / m * dZ2.dot(A1.T)\n",
    "    db2 = 1 / m * np.sum(dZ2)\n",
    "    dZ1 = W2.T.dot(dZ2) * ReLU_deriv(Z1)\n",
    "    dW1 = 1 / m * dZ1.dot(X.T)\n",
    "    db1 = 1 / m * np.sum(dZ1)\n",
    "    return dW1, db1, dW2, db2\n",
    "\n",
    "def update_params(W1, b1, W2, b2, dW1, db1, dW2, db2, alpha):\n",
    "    W1 = W1 - alpha * dW1\n",
    "    b1 = b1 - alpha * db1    \n",
    "    W2 = W2 - alpha * dW2  \n",
    "    b2 = b2 - alpha * db2    \n",
    "    return W1, b1, W2, b2"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3b9e3709-e669-4b0e-8956-cc7d2244a496",
   "metadata": {},
   "source": [
    "# Creating Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "a755a45b-e3aa-42b6-93f8-ca84fd73dac8",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_predictions(A2):\n",
    "    return np.argmax(A2, 0)\n",
    "\n",
    "def get_accuracy(predictions, Y):\n",
    "    print(predictions, Y)\n",
    "    return np.sum(predictions == Y) / Y.size\n",
    "\n",
    "def gradient_descent(X, Y, alpha, iterations):\n",
    "    W1, b1, W2, b2 = init_params()\n",
    "    for i in range(iterations):\n",
    "        Z1, A1, Z2, A2 = forward_prop(W1, b1, W2, b2, X)\n",
    "        dW1, db1, dW2, db2 = backward_prop(Z1, A1, Z2, A2, W1, W2, X, Y)\n",
    "        W1, b1, W2, b2 = update_params(W1, b1, W2, b2, dW1, db1, dW2, db2, alpha)\n",
    "        if i % 10 == 0:\n",
    "            print(\"Iteration: \", i)\n",
    "            predictions = get_predictions(A2)\n",
    "            print(get_accuracy(predictions, Y))\n",
    "    return W1, b1, W2, b2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "ee24ae06-b974-4ba7-8d86-49079f705c8a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration:  0\n",
      "[3 2 7 ... 7 8 6] [3 2 3 ... 7 0 2]\n",
      "0.11895121951219512\n",
      "Iteration:  10\n",
      "[3 2 7 ... 0 0 6] [3 2 3 ... 7 0 2]\n",
      "0.17673170731707316\n",
      "Iteration:  20\n",
      "[3 2 0 ... 0 0 6] [3 2 3 ... 7 0 2]\n",
      "0.26551219512195123\n",
      "Iteration:  30\n",
      "[3 2 0 ... 0 0 6] [3 2 3 ... 7 0 2]\n",
      "0.33151219512195124\n",
      "Iteration:  40\n",
      "[3 2 0 ... 0 0 6] [3 2 3 ... 7 0 2]\n",
      "0.36241463414634145\n",
      "Iteration:  50\n",
      "[3 2 0 ... 0 0 6] [3 2 3 ... 7 0 2]\n",
      "0.3958780487804878\n",
      "Iteration:  60\n",
      "[3 2 0 ... 0 0 6] [3 2 3 ... 7 0 2]\n",
      "0.4374390243902439\n",
      "Iteration:  70\n",
      "[3 2 0 ... 0 0 6] [3 2 3 ... 7 0 2]\n",
      "0.4809024390243902\n",
      "Iteration:  80\n",
      "[3 2 0 ... 0 0 3] [3 2 3 ... 7 0 2]\n",
      "0.522609756097561\n",
      "Iteration:  90\n",
      "[3 2 0 ... 0 0 3] [3 2 3 ... 7 0 2]\n",
      "0.5618048780487805\n",
      "Iteration:  100\n",
      "[3 2 0 ... 0 0 3] [3 2 3 ... 7 0 2]\n",
      "0.5970975609756097\n",
      "Iteration:  110\n",
      "[3 2 0 ... 0 0 3] [3 2 3 ... 7 0 2]\n",
      "0.6309756097560976\n",
      "Iteration:  120\n",
      "[3 2 0 ... 0 0 3] [3 2 3 ... 7 0 2]\n",
      "0.6576341463414634\n",
      "Iteration:  130\n",
      "[3 2 0 ... 0 0 3] [3 2 3 ... 7 0 2]\n",
      "0.6794146341463415\n",
      "Iteration:  140\n",
      "[3 2 0 ... 0 0 3] [3 2 3 ... 7 0 2]\n",
      "0.697780487804878\n",
      "Iteration:  150\n",
      "[3 2 0 ... 9 0 3] [3 2 3 ... 7 0 2]\n",
      "0.7141219512195122\n",
      "Iteration:  160\n",
      "[3 2 0 ... 9 0 3] [3 2 3 ... 7 0 2]\n",
      "0.7288292682926829\n",
      "Iteration:  170\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.7404146341463415\n",
      "Iteration:  180\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.7500243902439024\n",
      "Iteration:  190\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.7595609756097561\n",
      "Iteration:  200\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.7676585365853659\n",
      "Iteration:  210\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.7751707317073171\n",
      "Iteration:  220\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.781829268292683\n",
      "Iteration:  230\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.7875121951219513\n",
      "Iteration:  240\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.792609756097561\n",
      "Iteration:  250\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.7975121951219513\n",
      "Iteration:  260\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.8022682926829269\n",
      "Iteration:  270\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.8065609756097561\n",
      "Iteration:  280\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.8103658536585366\n",
      "Iteration:  290\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.8139024390243902\n",
      "Iteration:  300\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.817390243902439\n",
      "Iteration:  310\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.8207560975609756\n",
      "Iteration:  320\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.823780487804878\n",
      "Iteration:  330\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.8268292682926829\n",
      "Iteration:  340\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.8291951219512195\n",
      "Iteration:  350\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.8315853658536585\n",
      "Iteration:  360\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.8337317073170731\n",
      "Iteration:  370\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.8356097560975609\n",
      "Iteration:  380\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.8372682926829268\n",
      "Iteration:  390\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.839170731707317\n",
      "Iteration:  400\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.8411219512195122\n",
      "Iteration:  410\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.8423658536585366\n",
      "Iteration:  420\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.8437560975609756\n",
      "Iteration:  430\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.845170731707317\n",
      "Iteration:  440\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.8463658536585366\n",
      "Iteration:  450\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.8478292682926829\n",
      "Iteration:  460\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.8493170731707317\n",
      "Iteration:  470\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.851\n",
      "Iteration:  480\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.8520487804878049\n",
      "Iteration:  490\n",
      "[3 2 0 ... 7 0 3] [3 2 3 ... 7 0 2]\n",
      "0.8536585365853658\n"
     ]
    }
   ],
   "source": [
    "W1, b1, W2, b2 = gradient_descent(X_train, Y_train, 0.10, 500)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "413b4e93-2731-4c54-a761-152ebb05c2e6",
   "metadata": {},
   "source": [
    "# Testing Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "41c42d22-13c7-413b-8ea7-6ab46c580913",
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_predictions(X, W1, b1, W2, b2):\n",
    "    _, _, _, A2 = forward_prop(W1, b1, W2, b2, X)\n",
    "    predictions = get_predictions(A2)\n",
    "    return predictions\n",
    "\n",
    "def test_prediction(index, W1, b1, W2, b2):\n",
    "    current_image = X_train[:, index, None]\n",
    "    prediction = make_predictions(X_train[:, index, None], W1, b1, W2, b2)\n",
    "    label = Y_train[index]\n",
    "    print(\"Prediction: \", prediction)\n",
    "    print(\"Label: \", label)\n",
    "    \n",
    "    current_image = current_image.reshape((28, 28)) * 255\n",
    "    plt.gray()\n",
    "    plt.imshow(current_image, interpolation='nearest')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "1d4bf795-cb98-4c91-be4b-fadcd2a17531",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction:  [3]\n",
      "Label:  3\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction:  [2]\n",
      "Label:  2\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction:  [0]\n",
      "Label:  3\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction:  [8]\n",
      "Label:  8\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_prediction(0, W1, b1, W2, b2)\n",
    "test_prediction(1, W1, b1, W2, b2)\n",
    "test_prediction(2, W1, b1, W2, b2)\n",
    "test_prediction(3, W1, b1, W2, b2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "a1240197-55a2-449f-b99c-12e095a42c51",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[9 4 7 8 5 8 9 7 0 6 1 6 9 0 8 6 3 6 2 0 8 4 1 4 5 7 6 3 2 9 7 6 0 1 1 8 8\n",
      " 2 6 0 4 8 3 9 8 0 3 7 7 0 6 4 0 1 1 6 9 2 1 8 9 0 1 7 1 6 3 9 8 3 7 9 3 0\n",
      " 2 9 8 3 4 1 9 0 2 6 3 1 6 3 6 9 5 3 0 4 5 5 1 4 0 2 7 6 8 6 6 4 7 4 2 6 4\n",
      " 1 9 7 3 7 1 5 3 7 7 4 9 7 0 2 7 7 0 2 6 8 1 2 6 5 3 6 3 8 0 3 3 8 7 9 7 7\n",
      " 9 5 1 3 3 7 7 7 3 9 3 3 0 8 6 7 7 1 3 5 1 8 1 4 9 0 3 7 0 0 7 3 8 9 0 8 2\n",
      " 6 2 9 3 2 9 4 4 1 6 6 3 1 1 0 2 8 2 0 2 7 5 6 5 9 2 2 1 7 9 4 1 3 5 7 9 9\n",
      " 2 9 9 9 6 4 1 1 0 5 0 4 6 3 9 5 7 4 2 3 5 6 2 4 8 9 2 5 3 4 5 0 6 2 7 5 4\n",
      " 2 1 0 1 1 7 2 8 1 8 4 8 1 6 8 4 1 6 3 4 9 7 4 3 5 3 9 0 0 7 7 8 3 6 5 6 3\n",
      " 1 6 7 0 8 3 2 0 6 1 1 1 9 9 9 8 4 5 7 4 1 4 2 4 3 5 1 8 0 9 0 9 5 7 4 6 1\n",
      " 4 8 5 7 7 8 1 2 2 6 8 1 0 0 9 9 9 6 4 5 5 1 2 1 1 4 3 6 9 6 2 5 2 7 6 9 7\n",
      " 5 9 3 9 1 7 1 0 2 7 9 8 9 1 7 8 0 4 3 7 3 8 4 5 2 4 6 1 4 6 1 6 9 9 1 0 1\n",
      " 2 9 3 1 0 1 6 2 3 9 3 9 1 1 6 5 4 0 2 1 3 1 7 4 8 5 2 2 5 0 9 0 4 4 9 0 2\n",
      " 4 2 0 7 0 4 1 5 3 3 7 3 0 7 8 0 8 4 3 6 5 3 0 0 5 0 8 6 1 1 0 5 4 6 6 3 6\n",
      " 4 4 3 4 5 2 6 4 0 6 2 6 8 6 1 0 6 2 7 0 5 5 7 5 4 8 5 2 7 0 7 4 2 1 9 9 5\n",
      " 3 7 2 2 3 0 4 1 2 3 7 9 7 0 2 2 4 3 6 2 8 8 2 8 0 0 0 8 1 8 7 1 2 8 8 2 8\n",
      " 6 2 9 3 8 7 9 3 0 7 0 8 1 7 2 9 7 6 0 9 0 8 6 8 5 7 5 2 4 2 2 2 5 0 0 4 1\n",
      " 8 0 6 3 9 5 0 4 7 3 1 6 5 9 3 5 0 0 4 7 6 4 0 0 1 1 6 7 2 3 6 3 8 3 5 6 4\n",
      " 4 8 2 1 6 0 9 7 9 1 6 7 1 7 7 8 3 8 3 1 1 4 0 2 4 2 7 6 2 4 6 1 9 4 7 7 0\n",
      " 1 2 4 0 3 7 5 5 2 4 3 9 5 5 1 4 0 9 6 9 9 6 6 9 9 8 7 9 5 2 7 8 8 0 3 1 4\n",
      " 6 8 5 4 7 7 0 9 9 0 6 4 5 1 7 2 0 0 7 6 3 0 2 2 7 1 7 2 0 2 2 7 6 1 9 9 4\n",
      " 3 6 1 1 0 0 1 2 7 0 8 2 7 1 1 6 6 1 9 1 5 4 4 2 6 6 3 0 3 1 8 7 6 9 2 4 3\n",
      " 1 6 9 0 2 3 0 4 1 6 9 8 7 7 5 8 2 7 4 7 9 1 4 3 7 1 1 3 9 4 9 9 4 1 4 7 9\n",
      " 7 3 9 2 5 6 0 6 3 5 9 5 4 6 4 4 0 6 6 4 3 8 3 1 9 3 5 0 2 1 8 8 9 6 0 6 6\n",
      " 2 7 0 8 6 7 2 1 3 3 1 4 3 6 2 0 6 3 6 6 0 4 4 6 6 1 6 4 6 6 1 9 4 3 6 1 2\n",
      " 6 2 8 2 7 7 1 7 0 6 6 7 3 3 0 9 2 3 5 2 9 6 3 2 7 6 9 3 9 1 1 9 2 4 6 7 4\n",
      " 2 6 6 8 6 6 5 0 7 8 2 8 3 8 1 5 6 8 0 0 0 1 5 1 9 4 1 2 2 2 3 4 6 4 5 0 2\n",
      " 8 2 0 6 5 9 8 9 2 8 6 5 0 0 3 5 0 5 0 2 1 9 2 5 5 9 5 3 1 4 7 7 8 2 1 3 6\n",
      " 2] [9 4 7 8 5 2 9 8 0 6 1 6 9 0 8 6 3 0 2 0 5 4 1 9 5 7 6 3 2 9 7 0 0 1 1 8 8\n",
      " 2 6 0 4 8 3 9 8 0 3 7 7 0 6 4 0 1 1 6 9 2 1 8 9 0 1 9 1 6 3 4 8 5 7 9 5 0\n",
      " 2 7 8 3 2 1 9 0 2 6 3 8 6 3 6 7 5 3 0 4 5 5 1 4 7 2 7 6 8 6 6 4 7 4 2 6 4\n",
      " 1 7 7 3 7 1 0 8 7 7 4 9 7 0 2 7 7 0 2 6 8 1 2 6 5 3 4 3 5 0 3 3 8 3 4 3 9\n",
      " 9 5 1 3 3 7 2 7 3 9 3 3 0 8 6 7 7 1 3 5 1 8 5 9 0 0 3 7 0 0 7 3 8 9 0 5 2\n",
      " 6 2 9 3 8 9 2 4 1 6 6 3 1 1 5 2 8 2 0 2 7 8 6 5 9 2 2 1 7 9 4 1 3 0 6 9 9\n",
      " 2 9 9 9 6 4 1 1 0 5 0 4 6 3 7 5 7 4 2 3 5 6 2 6 8 8 2 3 3 4 5 0 6 2 7 5 4\n",
      " 2 1 0 1 1 7 7 8 1 8 4 8 1 6 8 4 7 5 3 4 9 7 4 3 5 3 9 0 0 7 7 8 3 6 5 6 3\n",
      " 1 6 7 0 8 3 2 0 6 1 1 1 9 9 9 8 4 5 7 4 1 9 8 4 3 5 7 3 0 9 3 9 8 7 4 6 1\n",
      " 4 8 5 7 7 3 1 2 2 6 5 1 0 0 9 9 9 0 4 5 5 1 2 1 1 4 3 6 9 6 2 5 3 7 6 9 7\n",
      " 5 9 3 9 6 7 1 0 2 7 9 8 9 1 7 8 0 4 3 7 3 8 4 2 2 4 6 1 4 6 1 6 9 4 1 0 1\n",
      " 2 9 3 1 0 1 6 2 3 9 8 9 1 1 6 5 4 0 2 1 2 7 7 4 7 5 2 2 5 0 9 0 9 4 9 0 2\n",
      " 4 2 0 7 7 4 2 5 3 3 7 3 0 7 8 0 8 4 3 4 5 5 0 0 5 0 8 6 1 1 0 5 4 6 6 8 6\n",
      " 4 4 3 4 5 2 2 4 0 6 2 6 8 6 1 0 6 6 7 0 3 8 7 0 9 8 0 2 7 0 5 4 2 1 9 7 5\n",
      " 3 7 2 2 3 0 4 1 2 3 2 0 7 0 2 2 4 2 6 2 8 8 2 8 0 5 8 8 1 8 7 1 2 8 3 2 8\n",
      " 6 3 9 3 8 7 9 3 0 7 0 8 7 7 2 3 7 6 0 9 0 8 6 8 5 7 5 2 4 2 2 5 5 0 0 4 1\n",
      " 8 0 6 3 5 5 0 4 7 3 7 6 5 9 3 5 0 5 4 7 6 4 0 0 1 1 6 7 2 3 6 3 8 0 5 6 4\n",
      " 4 8 2 1 6 0 9 7 9 9 6 7 1 7 7 8 3 5 3 1 2 4 0 2 6 2 7 6 2 9 6 1 5 4 9 7 0\n",
      " 1 2 4 0 3 7 5 5 4 4 3 9 5 5 1 4 0 9 6 9 9 6 6 8 9 8 7 9 5 2 7 8 3 0 3 1 4\n",
      " 6 8 5 2 7 7 0 9 9 0 2 4 5 1 7 2 0 0 7 6 3 0 2 1 7 1 7 2 0 2 2 7 6 1 9 9 4\n",
      " 3 6 1 1 0 0 1 2 7 0 8 2 9 1 1 6 6 1 9 1 5 4 6 2 6 6 3 0 5 1 8 7 6 9 2 4 3\n",
      " 1 6 9 0 2 3 0 4 1 6 2 8 7 7 6 5 2 7 4 7 9 1 4 5 7 1 1 2 9 4 9 9 4 1 4 7 9\n",
      " 7 3 9 2 5 6 0 6 3 5 9 5 9 6 4 4 6 6 6 4 3 8 3 1 9 2 1 7 2 1 8 4 9 6 0 6 6\n",
      " 2 7 0 8 6 7 2 1 3 3 1 4 3 6 2 0 4 3 6 6 0 4 9 0 6 1 6 4 6 6 1 9 4 3 4 1 2\n",
      " 6 2 5 2 7 7 1 7 5 6 6 7 3 3 0 9 2 3 5 2 9 6 3 2 7 6 4 3 9 1 1 9 7 4 6 7 4\n",
      " 0 6 5 8 6 6 5 0 7 8 2 8 3 2 1 3 6 8 0 0 0 1 5 1 9 9 1 6 2 2 3 4 6 4 5 0 2\n",
      " 8 2 0 6 5 9 5 9 2 8 6 5 0 0 5 5 0 5 0 2 1 9 2 5 5 7 5 3 2 2 7 7 8 5 1 3 6\n",
      " 2]\n",
      "0.865\n"
     ]
    }
   ],
   "source": [
    "dev_predictions = make_predictions(X_dev, W1, b1, W2, b2)\n",
    "print(get_accuracy(dev_predictions, Y_dev))"
   ]
  }
 ],
 "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": "3.10.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}