{ "cells": [ { "cell_type": "code", "execution_count": 1, "id": "bee88c73-8bd2-4d3c-ae56-c289c8262a48", "metadata": {}, "outputs": [], "source": [ "import torch\n", "import torch.nn as nn\n", "import torch.optim as optim\n", "from torch.utils.data import TensorDataset, DataLoader\n", "import pandas as pd" ] }, { "cell_type": "markdown", "id": "500c43c0-2cd6-44f1-b189-c69646b3ceeb", "metadata": {}, "source": [ "# Dataset" ] }, { "cell_type": "code", "execution_count": 2, "id": "9616b609-a2cc-405d-a23f-84723402fb1d", "metadata": {}, "outputs": [], "source": [ "# Load to Pandas Dataframe\n", "train_path = 'training.dat'\n", "test_path = 'testing.dat'\n", "\n", "train_df = pd.read_csv(train_path, delim_whitespace=True, header=None)\n", "test_df = pd.read_csv(test_path, delim_whitespace=True, header=None)\n", "\n", "# Assuming that all columns are features and classes are based on row indices\n", "X_train = torch.tensor(train_df.values).float() # Convert feature data to tensor\n", "X_test = torch.tensor(test_df.values).float() # Convert feature data to tensor\n", "\n", "# Create labels based on row indices\n", "y_train = torch.tensor([(i // 25) for i in range(len(train_df))]).long()\n", "y_test = torch.tensor([(i // 25) for i in range(len(test_df))]).long()\n", "\n", "# Create data loaders\n", "train_loader = DataLoader(TensorDataset(X_train, y_train), batch_size=5, shuffle=True)\n", "test_loader = DataLoader(TensorDataset(X_test, y_test), batch_size=5)" ] }, { "cell_type": "markdown", "id": "9367529a-9ce1-4aed-ae7a-9ad49b0870bb", "metadata": {}, "source": [ "# Model" ] }, { "cell_type": "code", "execution_count": 4, "id": "4723d5ce-2f2d-4dc9-b4a4-40639588b152", "metadata": {}, "outputs": [], "source": [ "class NeuralNetwork(nn.Module):\n", " def __init__(self):\n", " super(NeuralNetwork, self).__init__()\n", " self.hidden1 = nn.Linear(4, 8)\n", " self.hidden2 = nn.Linear(8, 8)\n", " self.hidden3 = nn.Linear(8, 8)\n", " self.hidden4 = nn.Linear(8, 8)\n", " self.output = nn.Linear(8, 3)\n", " self.relu = nn.ReLU()\n", "\n", " def forward(self, x):\n", " x = self.relu(self.hidden1(x))\n", " x = self.relu(self.hidden2(x))\n", " x = self.relu(self.hidden3(x))\n", " x = self.relu(self.hidden4(x))\n", " x = self.output(x)\n", " return x" ] }, { "cell_type": "markdown", "id": "b08bc191-9c9c-423c-8e3c-597099f2a519", "metadata": {}, "source": [ "# Train" ] }, { "cell_type": "code", "execution_count": 5, "id": "f7f3d240-0ba4-443e-9b1c-172419c311d3", "metadata": {}, "outputs": [], "source": [ "# Instantiate the model, loss function, and optimizer\n", "model = NeuralNetwork()\n", "criterion = nn.CrossEntropyLoss()\n", "optimizer = optim.Adam(model.parameters(), lr=0.001)" ] }, { "cell_type": "code", "execution_count": 6, "id": "56b466a1-cdaa-4143-b097-6737a91eeed3", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Epoch 1, Loss: 1.1656708717346191\n", "Epoch 2, Loss: 1.069056749343872\n", "Epoch 3, Loss: 1.1333191394805908\n", "Epoch 4, Loss: 1.1097159385681152\n", "Epoch 5, Loss: 1.0479941368103027\n", "Epoch 6, Loss: 1.1416243314743042\n", "Epoch 7, Loss: 1.016518235206604\n", "Epoch 8, Loss: 1.1750520467758179\n", "Epoch 9, Loss: 1.0874340534210205\n", "Epoch 10, Loss: 1.0893971920013428\n", "Epoch 11, Loss: 1.0989606380462646\n", "Epoch 12, Loss: 1.1664495468139648\n", "Epoch 13, Loss: 1.0913177728652954\n", "Epoch 14, Loss: 1.0568711757659912\n", "Epoch 15, Loss: 0.9801815152168274\n", "Epoch 16, Loss: 0.9334077835083008\n", "Epoch 17, Loss: 0.7538875341415405\n", "Epoch 18, Loss: 0.7881473302841187\n", "Epoch 19, Loss: 0.7299363613128662\n", "Epoch 20, Loss: 0.2598203122615814\n", "Epoch 21, Loss: 0.7928000688552856\n", "Epoch 22, Loss: 0.7305868864059448\n", "Epoch 23, Loss: 0.2939522862434387\n", "Epoch 24, Loss: 0.43748903274536133\n", "Epoch 25, Loss: 0.25262290239334106\n", "Epoch 26, Loss: 0.2567386031150818\n", "Epoch 27, Loss: 0.32417288422584534\n", "Epoch 28, Loss: 0.36710914969444275\n", "Epoch 29, Loss: 0.44611167907714844\n", "Epoch 30, Loss: 0.2571166157722473\n", "Epoch 31, Loss: 0.3475134074687958\n", "Epoch 32, Loss: 0.33021003007888794\n", "Epoch 33, Loss: 0.24303965270519257\n", "Epoch 34, Loss: 0.14755365252494812\n", "Epoch 35, Loss: 0.07536564767360687\n", "Epoch 36, Loss: 0.16837753355503082\n", "Epoch 37, Loss: 0.1387643814086914\n", "Epoch 38, Loss: 0.44958290457725525\n", "Epoch 39, Loss: 0.09252478182315826\n", "Epoch 40, Loss: 0.10783408582210541\n", "Epoch 41, Loss: 0.031535036861896515\n", "Epoch 42, Loss: 0.06791599839925766\n", "Epoch 43, Loss: 0.05487314984202385\n", "Epoch 44, Loss: 0.07081137597560883\n", "Epoch 45, Loss: 0.039346300065517426\n", "Epoch 46, Loss: 0.02455393224954605\n", "Epoch 47, Loss: 0.04027646780014038\n", "Epoch 48, Loss: 0.22815446555614471\n", "Epoch 49, Loss: 0.006961571518331766\n", "Epoch 50, Loss: 0.04802575707435608\n" ] } ], "source": [ "# Training loop\n", "for epoch in range(50): # number of epochs\n", " for inputs, labels in train_loader:\n", " optimizer.zero_grad()\n", " outputs = model(inputs)\n", " loss = criterion(outputs, labels) # 오차(오류) 계산\n", " loss.backward() # 오차(오류) 역전파\n", " optimizer.step() # 모델 가중치값 업데이트 -> Adam optimizer 기반\n", "\n", " print(f'Epoch {epoch+1}, Loss: {loss.item()}')" ] }, { "cell_type": "markdown", "id": "faad1d77-d622-4d15-9a8a-82d5af1cfdec", "metadata": {}, "source": [ "# Test" ] }, { "cell_type": "code", "execution_count": 7, "id": "765d9b68-6e8b-4c34-a03c-c56356fed22f", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Accuracy: 94.66666666666667%\n" ] } ], "source": [ "# Testing the model\n", "correct = 0\n", "total = 0\n", "with torch.no_grad():\n", " for inputs, labels in test_loader:\n", " outputs = model(inputs)\n", " _, predicted = torch.max(outputs.data, 1)\n", " total += labels.size(0)\n", " correct += (predicted == labels).sum().item()\n", "\n", "print(f'Accuracy: {100 * correct / total}%')" ] } ], "metadata": { "kernelspec": { "display_name": "prac", "language": "python", "name": "prac" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.9.0" } }, "nbformat": 4, "nbformat_minor": 5 }