Anu Swathi - Task4

Logistic Regression for Binary Classifier

This project demonstrates the implementation of a binary classification model using Logistic Regression on the Breast Cancer Wisconsin Diagnostic Dataset.

🧠 Objective

Build a binary classifier using logistic regression and evaluate its performance using various metrics and visualization tools.

🛠 Tools Used

• Python
• Scikit-learn
• Pandas
• Matplotlib
• Seaborn

📁 Dataset

The dataset used is the Breast Cancer Wisconsin Diagnostic dataset, which contains features computed from digitized images of fine needle aspirates (FNA) of breast masses.

🧪 Steps Covered

1. Data Loading & Exploration
2. Preprocessing
   (i) Dropping unnecessary columns
   (ii) Converting categorical labels to binary
3. Train/Test Split and Feature Standardization
4. Model Training with Logistic Regression
5. Model Evaluation
   (i) Confusion Matrix
   (ii) Precision, Recall, F1-score
   (iii) ROC-AUC score
   (iv) ROC Curve Visualization
6. Threshold Tuning
   Precision and recall at different thresholds
7. Sigmoid Function Explanation

📊 Results

1. Achieved high accuracy and AUC using Logistic Regression.

2. Tuned threshold to observe trade-offs between precision and recall.