ia/ml-labs/Laboratorul_7_Perceptron.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Laboratorul 7 Perceptron",
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "xR7wegrOus-t",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "t1bmFfIMudR-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X = np.array([[0, 0], [0, 1], [1, 0], [1, 1]])\n",
        "y = np.array([-1, 1, 1, 1])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mRPgblHluwPl",
        "colab_type": "code",
        "outputId": "0150d11d-b069-4457-d274-c6de56596200",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        }
      },
      "source": [
        "plt.scatter(X[:, 0], X[:, 1], c=y, cmap='seismic')\n",
        "plt.plot([-0.1, 0.8], [0.8, -0.1])\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "U_RhH-1GvBja",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def compute_y(x, W, bias):\n",
        "    # dreapta de decizie\n",
        "    # [x, y] * [W[0], W[1]] + b = 0\n",
        "    EPSILON = 1e-10\n",
        "    return (-x * W[0] - bias) / (W[1] + EPSILON)\n",
        "\n",
        "def plot_decision_boundary(X, y, W, b, current_x, current_y):\n",
        "    x1 = -0.5\n",
        "    y1 = compute_y(x1, W, b)\n",
        "    x2 = 0.5\n",
        "    y2 = compute_y(x2, W, b)\n",
        "    \n",
        "    # sterge continutul ferestrei\n",
        "\n",
        "    # ploteaza multimea de antrenare\n",
        "    color = 'r'\n",
        "    if current_y == -1:\n",
        "        color = 'b'\n",
        "    plt.ylim((-1, 2))\n",
        "    plt.xlim((-1, 2))\n",
        "    plt.plot(X[y == -1, 0], X[y == -1, 1], 'b+')\n",
        "    plt.plot(X[y == 1, 0], X[y == 1, 1], 'r+')\n",
        "\n",
        "    # ploteaza exemplul curent\n",
        "    plt.plot(current_x[0], current_x[1], color+'s')\n",
        "\n",
        "    # afisarea dreptei de decizie\n",
        "    plt.plot([x1, x2] ,[y1, y2], 'black')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aa4Mi0cpwQec",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "W = np.random.randn(2)\n",
        "b = np.random.randn(1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Z7bag6Ijw41J",
        "colab_type": "code",
        "outputId": "0371212a-1f06-4416-bcbf-6c68fd6d6261",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 269
        }
      },
      "source": [
        "plot_decision_boundary(X, y, W, b, X[0], y[0])\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KXOiF7uAxKhj",
        "colab_type": "code",
        "outputId": "ed4c2e25-4b91-4aeb-8292-2d3e537194a9",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        }
      },
      "source": [
        "epochs = 70\n",
        "learning_rate = 1e-1\n",
        "\n",
        "total_loss = 0\n",
        "total_count = 0\n",
        "for epoch in range(epochs):\n",
        "    for x, y_true in zip(X, y):\n",
        "        y_hat = x @ W + b\n",
        "        y_diff = y_hat - y_true\n",
        "        loss = (y_diff ** 2) / 2\n",
        "        total_loss += loss.item()\n",
        "        W -= learning_rate * y_diff * x\n",
        "        b -= learning_rate * y_diff\n",
        "    total_count += len(y)\n",
        "\n",
        "average_loss = total_loss / total_count\n",
        "print(average_loss)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.34437283099315025\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nbtwkWoe0auu",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "num_hidden_neurons = 3\n",
        "\n",
        "W_1 = np.random.randn(2, num_hidden_neurons)\n",
        "b_1 = np.zeros(num_hidden_neurons)\n",
        "\n",
        "W_2 = np.random.randn(num_hidden_neurons, 1)\n",
        "b_2 = np.zeros(1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "u6sdSxJd3OV2",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def sigmoid(z):\n",
        "    return 1 / (1 + np.exp(-z))\n",
        "\n",
        "def forward(X):\n",
        "    z_1 = X @ W_1 + b_1\n",
        "    a_1 = np.tanh(z_1)\n",
        "    z_2 = a_1 @ W_2 + b_2\n",
        "    a_2 = sigmoid(z_2)\n",
        "    return z_1, a_1, z_2, a_2\n",
        "\n",
        "z_1, a_1, z_2, a_2 = forward(X)\n",
        "\n",
        "\n",
        "loss = (-y * np.log(a_2) - (1 - y) * np.log(1 - a_2)).mean()\n",
        "accuracy = (np.round(a_2) == y).mean()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iKe6GJjD31RB",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def backward(a_1, a_2, z_1, W_2, X, Y, num_samples):\n",
        "    dz_2 = a_2 - y\n",
        "    dw_2 = (a_1.T @ d_z2) / num_samples\n",
        "    db_2 = np.sum(dz_2) / num_samples\n",
        "    da_1 = dz_2 @ W_2.T\n",
        "\n",
        "    dz_1 = da_1 * tanh_derivative(z_1)\n",
        "\n",
        "    dw_1 = X.T @ dz_1 / num_samples\n",
        "\n",
        "    db_1 = np.sum(dz_1) / num_samples\n",
        "\n",
        "    return dw_1, db_1, dw_2, db_2"
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}