course_work_final.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/DenisB22/Mushroom-Classification-ML/blob/main/course_work_final.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "import torch.optim as optim\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "from sklearn.preprocessing import LabelEncoder\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "from imblearn.over_sampling import RandomOverSampler\n",
        "\n",
        "# Define the neural network model\n",
        "class Net(nn.Module):\n",
        "    def __init__(self):\n",
        "        super(Net, self).__init__()\n",
        "        self.fc1 = nn.Linear(128, 64)\n",
        "        self.fc2 = nn.Linear(64, 32)\n",
        "        self.fc3 = nn.Linear(32, 1)\n",
        "\n",
        " \n",
        "\n",
        "    def forward(self, x):\n",
        "        x = F.relu(self.fc1(x))\n",
        "        x = F.relu(self.fc2(x))\n",
        "        \n",
        "        x = torch.sigmoid(self.fc3(x))\n",
        "        \n",
        "        return x\n",
        "\n",
        "# Read and preprocess the data\n",
        "train_data = pd.read_csv(\"/content/shroom/train_data.csv\")\n",
        "test_data = pd.read_csv(\"/content/shroom/test_data.csv\")\n",
        "train_data = train_data.drop(columns=[\"id\"]) # Drop the ID column\n",
        "test_data_ids = test_data[\"id\"].values # Get the test data IDs\n",
        "test_data = test_data.drop(columns=[\"id\"]) # Drop the ID column\n",
        "\n",
        "train_numeric_features = train_data.dtypes[train_data.dtypes != 'object'].index\n",
        "test_numeric_features = test_data.dtypes[test_data.dtypes != 'object'].index\n",
        "\n",
        "train_data[train_numeric_features] = train_data[train_numeric_features].apply(\n",
        "    lambda x: (x - x.mean()) / (x.std()))\n",
        "\n",
        "test_data[test_numeric_features] = test_data[test_numeric_features].apply(\n",
        "    lambda x: (x - x.mean()) / (x.std()))\n",
        "\n",
        "train_data[train_numeric_features] = train_data[train_numeric_features].fillna(0)\n",
        "test_data[test_numeric_features] = test_data[test_numeric_features].fillna(0)\n",
        "train_data[\"poisonous\"] = train_data[\"poisonous\"].apply(lambda x: 1 if x == 'y' else 0)\n",
        "\n",
        "train_data = pd.get_dummies(train_data, dummy_na=True)\n",
        "test_data = pd.get_dummies(test_data, dummy_na=True)\n",
        "\n",
        "# print(train_data['poisonous_y'])\n",
        "\n",
        "# le = LabelEncoder() # Encode categorical data as integers\n",
        "# train_data = train_data.apply(le.fit_transform)\n",
        "# # test_data = test_data.fillna(test_data.mode().iloc[0])\n",
        "# test_data = test_data.apply(le.fit_transform)\n",
        "\n",
        "X_train = train_data.drop(columns=[\"poisonous\"]).values\n",
        "y_train = train_data[\"poisonous\"].values\n",
        "X_test = test_data.values\n",
        "\n",
        "oversample = RandomOverSampler(random_state=42)\n",
        "X_train, y_train = oversample.fit_resample(X_train, y_train)\n",
        "\n",
        "X_train, X_val, y_train, y_val = train_test_split(X_train, y_train, test_size=0.2)\n",
        "\n",
        "X_train = torch.from_numpy(X_train).float()\n",
        "y_train = torch.from_numpy(y_train).float().view(-1, 1)\n",
        "\n",
        "X_val = torch.from_numpy(X_val).float()\n",
        "y_val = torch.from_numpy(y_val).float().view(-1, 1)\n",
        "\n",
        "X_test = torch.from_numpy(X_test).float()\n",
        "\n",
        "# Train the neural network model\n",
        "net = Net()\n",
        "criterion = nn.BCELoss()\n",
        "# optimizer = optim.SGD(net.parameters(), lr=0.01)\n",
        "optimizer = optim.Adam(net.parameters(), lr=0.01)\n",
        "train_ls = []\n",
        "\n",
        "for epoch in range(1200):\n",
        "    optimizer.zero_grad()\n",
        "    output = net(X_train)\n",
        "    loss = criterion(output, y_train)\n",
        "    loss.backward()\n",
        "    optimizer.step()\n",
        "    train_ls.append(loss.item())\n",
        "    if epoch % 100 == 0:\n",
        "        print(f\"Epoch {epoch+1}/{1000}, Loss: {loss.item()}\")\n",
        "\n",
        "# Test the neural network model\n",
        "with torch.no_grad():\n",
        "    net.eval()\n",
        "    output = net(X_val)\n",
        "    predicted = (output >= 0.5).float()\n",
        "    accuracy = (predicted == y_val).float().mean()\n",
        "    print(f\"Validation Accuracy: {accuracy.item()}\")\n",
        "\n",
        "# Generate predictions for the test data\n",
        "with torch.no_grad():\n",
        "    net.eval()\n",
        "    output = net(X_test)\n",
        "    predicted = (output >= 0.5).float().view(-1)\n",
        "    predicted_labels = l.inverse_transform(predicted.int().numpy())\n",
        "\n",
        "# Write the predicted labels to a CSV file\n",
        "output_df = pd.DataFrame({\"id\": test_data_ids, \"poisonous\": predicted_labels})\n",
        "output_df.to_csv(\"result.csv\", index=False)\n",
        "\n",
        "# Load the predicted labels from the result.csv file\n",
        "df = pd.read_csv(\"/content/result.csv\")\n",
        "\n",
        "# Convert the numerical labels to \"y\" or \"n\"\n",
        "df[\"poisonous\"] = df[\"poisonous\"].apply(lambda x: \"y\" if x == 's' else \"n\")\n",
        "\n",
        "# Save the updated labels to a new file\n",
        "df.to_csv(\"result_updated.csv\", index=False)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 460
        },
        "id": "077wgiTdMEmu",
        "outputId": "6d74c04e-9683-4d95-8683-3f50a3684364"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/1000, Loss: 0.6937305331230164\n",
            "Epoch 101/1000, Loss: 0.00019353243988007307\n",
            "Epoch 201/1000, Loss: 8.46208495204337e-05\n",
            "Epoch 301/1000, Loss: 4.832736885873601e-05\n",
            "Epoch 401/1000, Loss: 3.1347357435151935e-05\n",
            "Epoch 501/1000, Loss: 2.196307286794763e-05\n",
            "Epoch 601/1000, Loss: 1.6230724213528447e-05\n",
            "Epoch 701/1000, Loss: 1.2485690604080446e-05\n",
            "Epoch 801/1000, Loss: 9.897690688376315e-06\n",
            "Epoch 901/1000, Loss: 8.025695933611132e-06\n",
            "Epoch 1001/1000, Loss: 6.6259599407203496e-06\n",
            "Epoch 1101/1000, Loss: 5.557154509006068e-06\n",
            "Validation Accuracy: 0.9955357313156128\n"
          ]
        },
        {
          "output_type": "error",
          "ename": "NameError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-2-24079d06c555>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m    105\u001b[0m     \u001b[0moutput\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnet\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX_test\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    106\u001b[0m     \u001b[0mpredicted\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0moutput\u001b[0m \u001b[0;34m>=\u001b[0m \u001b[0;36m0.5\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfloat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mview\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 107\u001b[0;31m     \u001b[0mpredicted_labels\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mle\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0minverse_transform\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpredicted\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnumpy\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    108\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    109\u001b[0m \u001b[0;31m# Write the predicted labels to a CSV file\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mNameError\u001b[0m: name 'le' is not defined"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "plt.plot(np.arange(1, 1200+1), train_ls)\n",
        "plt.xlabel(\"epoch\")\n",
        "plt.ylabel(\"loss\")\n",
        "plt.yscale(\"log\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 281
        },
        "id": "ShiyvMkROBbC",
        "outputId": "cf2819f2-2d31-45e5-be49-b120238d9507"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}