{ "nbformat": 4, "nbformat_minor": 0, "metadata": { "colab": { "provenance": [] }, "kernelspec": { "name": "python3", "display_name": "Python 3" }, "language_info": { "name": "python" } }, "cells": [ { "cell_type": "code", "source": [ "import numpy as np\n", "import pandas as pd\n", "import matplotlib.pyplot as plt\n", "from sklearn.model_selection import train_test_split\n", "\n", "data = pd.read_csv(\"data.csv\")\n", "print (data.head)\n", "data.info()\n", "\n", "\n", "# Preprocess data\n", "data.drop(['Unnamed: 32', 'id'], axis=1, inplace=True)\n", "data['diagnosis'] = data['diagnosis'].map({'M': 1, 'B': 0})\n", "\n", "y = data['diagnosis'].values\n", "x_data = data.drop(['diagnosis'], axis=1)\n", "\n", "# Normalize features\n", "x = (x_data - x_data.min()) / (x_data.max() - x_data.min())\n", "\n", "from sklearn.model_selection import train_test_split\n", "x_train, x_test, y_train, y_test = train_test_split(\n", " x, y, test_size = 0.15, random_state = 42)\n", "\n", "x_train = x_train.T\n", "x_test = x_test.T\n", "y_train = y_train.T\n", "y_test = y_test.T\n", "\n", "print(\"x train: \", x_train.shape)\n", "print(\"x test: \", x_test.shape)\n", "print(\"y train: \", y_train.shape)\n", "print(\"y test: \", y_test.shape)\n", "\n", "\n", "def initialize_weights_and_bias(dimension):\n", " w = np.random.randn(dimension, 1) * 0.01 # Initialize with small random values\n", " b = 0.0\n", " return w, b\n", "\n", "def sigmoid(z):\n", " return 1 / (1 + np.exp(-z))\n", "\n", "def forward_backward_propagation(w, b, x_train, y_train):\n", " m = x_train.shape[1]\n", " z = np.dot(w.T, x_train) + b\n", " y_head = sigmoid(z)\n", "\n", " # Calculate cost\n", " cost = (-1/m) * np.sum(y_train * np.log(y_head) + (1 - y_train) * np.log(1 - y_head))\n", "\n", " # Backward propagation\n", " derivative_weight = (1/m) * np.dot(x_train, (y_head - y_train).T)\n", " derivative_bias = (1/m) * np.sum(y_head - y_train)\n", "\n", " gradients = {\"derivative_weight\": derivative_weight, \"derivative_bias\": derivative_bias}\n", " return cost, gradients\n", "\n", "def update(w, b, x_train, y_train, learning_rate, num_iterations):\n", " costs = []\n", " gradients = {} # Store gradients\n", " for i in range(num_iterations):\n", " cost, grad = forward_backward_propagation(w, b, x_train, y_train)\n", " w -= learning_rate * grad[\"derivative_weight\"]\n", " b -= learning_rate * grad[\"derivative_bias\"]\n", "\n", " if i % 100 == 0:\n", " costs.append(cost)\n", " print(f\"Cost after iteration {i}: {cost}\")\n", "\n", " parameters = {\"weight\": w, \"bias\": b}\n", " return parameters, gradients, costs\n", "\n", "\n", "def predict(w, b, x_test):\n", " m = x_test.shape[1]\n", " y_prediction = np.zeros((1, m))\n", " z = sigmoid(np.dot(w.T, x_test) + b)\n", "\n", " for i in range(z.shape[1]):\n", " y_prediction[0, i] = 1 if z[0, i] > 0.5 else 0\n", "\n", " return y_prediction\n", "\n", "def logistic_regression(x_train, y_train, x_test, y_test, learning_rate=0.01, num_iterations=1000):\n", " dimension = x_train.shape[0]\n", " w, b = initialize_weights_and_bias(dimension)\n", " parameters, gradients, costs = update(w, b, x_train, y_train, learning_rate, num_iterations)\n", "\n", " y_prediction_test = predict(parameters[\"weight\"], parameters[\"bias\"], x_test)\n", " y_prediction_train = predict(parameters[\"weight\"], parameters[\"bias\"], x_train)\n", "\n", " print(f\"Train accuracy: {100 - np.mean(np.abs(y_prediction_train - y_train)) * 100}%\")\n", " print(f\"Test accuracy: {100 - np.mean(np.abs(y_prediction_test - y_test)) * 100}%\")\n", "\n", "# Run logistic regression\n", "logistic_regression(x_train, y_train, x_test, y_test, learning_rate=0.01, num_iterations=1000)" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "7_PZ6lqtX3pB", "outputId": "4b6f8628-2896-4253-a2c2-26a43cb02dcb" }, "execution_count": 4, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "\n", "\n", "RangeIndex: 569 entries, 0 to 568\n", "Data columns (total 33 columns):\n", " # Column Non-Null Count Dtype \n", "--- ------ -------------- ----- \n", " 0 id 569 non-null int64 \n", " 1 diagnosis 569 non-null object \n", " 2 radius_mean 569 non-null float64\n", " 3 texture_mean 569 non-null float64\n", " 4 perimeter_mean 569 non-null float64\n", " 5 area_mean 569 non-null float64\n", " 6 smoothness_mean 569 non-null float64\n", " 7 compactness_mean 569 non-null float64\n", " 8 concavity_mean 569 non-null float64\n", " 9 concave points_mean 569 non-null float64\n", " 10 symmetry_mean 569 non-null float64\n", " 11 fractal_dimension_mean 569 non-null float64\n", " 12 radius_se 569 non-null float64\n", " 13 texture_se 569 non-null float64\n", " 14 perimeter_se 569 non-null float64\n", " 15 area_se 569 non-null float64\n", " 16 smoothness_se 569 non-null float64\n", " 17 compactness_se 569 non-null float64\n", " 18 concavity_se 569 non-null float64\n", " 19 concave points_se 569 non-null float64\n", " 20 symmetry_se 569 non-null float64\n", " 21 fractal_dimension_se 569 non-null float64\n", " 22 radius_worst 569 non-null float64\n", " 23 texture_worst 569 non-null float64\n", " 24 perimeter_worst 569 non-null float64\n", " 25 area_worst 569 non-null float64\n", " 26 smoothness_worst 569 non-null float64\n", " 27 compactness_worst 569 non-null float64\n", " 28 concavity_worst 569 non-null float64\n", " 29 concave points_worst 569 non-null float64\n", " 30 symmetry_worst 569 non-null float64\n", " 31 fractal_dimension_worst 569 non-null float64\n", " 32 Unnamed: 32 0 non-null float64\n", "dtypes: float64(31), int64(1), object(1)\n", "memory usage: 146.8+ KB\n", "x train: (30, 483)\n", "x test: (30, 86)\n", "y train: (483,)\n", "y test: (86,)\n", "Cost after iteration 0: 0.6928136132424735\n", "Cost after iteration 100: 0.6650819203054067\n", "Cost after iteration 200: 0.6406057807006481\n", "Cost after iteration 300: 0.6182622214830192\n", "Cost after iteration 400: 0.5976893069647797\n", "Cost after iteration 500: 0.5786892834063715\n", "Cost after iteration 600: 0.5611076230947329\n", "Cost after iteration 700: 0.5448087404195209\n", "Cost after iteration 800: 0.529670875172974\n", "Cost after iteration 900: 0.5155845404270566\n", "Train accuracy: 90.6832298136646%\n", "Test accuracy: 88.37209302325581%\n" ] } ] }, { "cell_type": "code", "source": [], "metadata": { "id": "sWcy1yQqaVC3" }, "execution_count": 4, "outputs": [] }, { "cell_type": "code", "source": [], "metadata": { "id": "GAjYhEzbjWli" }, "execution_count": null, "outputs": [] } ] }