Computer_Vision/Chapter01/Gradient_descent.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Gradient_descent.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "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": "3.7.6"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/PacktPublishing/Modern-Computer-Vision-with-PyTorch/blob/master/Chapter01/Gradient_descent.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "ExecuteTime": {
          "end_time": "2020-09-24T13:36:49.845714Z",
          "start_time": "2020-09-24T13:36:49.666592Z"
        },
        "id": "TTWK_bQBC7mf"
      },
      "source": [
        "import numpy as np \n",
        "from copy import deepcopy\n",
        "import matplotlib.pyplot as plt\n",
        "x = np.array([[1,1]])\n",
        "y = np.array([[0]])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "ExecuteTime": {
          "end_time": "2020-09-24T13:36:49.850377Z",
          "start_time": "2020-09-24T13:36:49.846832Z"
        },
        "id": "V8f13jAKC9Rv"
      },
      "source": [
        "from copy import deepcopy\n",
        "import numpy as np\n",
        "def feed_forward(inputs, outputs, weights):     \n",
        "    pre_hidden = np.dot(inputs,weights[0])+ weights[1]\n",
        "    hidden = 1/(1+np.exp(-pre_hidden))\n",
        "    out = np.dot(hidden, weights[2]) + weights[3]\n",
        "    mean_squared_error = np.mean(np.square(out - outputs))\n",
        "    return mean_squared_error"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "ExecuteTime": {
          "end_time": "2020-09-24T13:36:49.855575Z",
          "start_time": "2020-09-24T13:36:49.851797Z"
        },
        "id": "8mF3nQbMC_Ja"
      },
      "source": [
        "def update_weights(inputs, outputs, weights, lr):\n",
        "    original_weights = deepcopy(weights)\n",
        "    temp_weights = deepcopy(weights)\n",
        "    updated_weights = deepcopy(weights)\n",
        "    original_loss = feed_forward(inputs, outputs, original_weights)\n",
        "    for i, layer in enumerate(original_weights):\n",
        "        for index, weight in np.ndenumerate(layer):\n",
        "            temp_weights = deepcopy(weights)\n",
        "            temp_weights[i][index] += 0.0001\n",
        "            _loss_plus = feed_forward(inputs, outputs, temp_weights)\n",
        "            grad = (_loss_plus - original_loss)/(0.0001)\n",
        "            updated_weights[i][index] -= grad*lr\n",
        "    return updated_weights, original_loss"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QqapuMCwDQiJ"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}