Breast Cancer Prediction with Machine Learning

This project demonstrates an end-to-end machine learning pipeline for predicting breast cancer using the Breast Cancer Wisconsin Diagnostic Dataset. The Logistic Regression model classifies tumors as malignant (cancerous) or benign (non-cancerous), showcasing the full ML workflow from data cleaning to model evaluation.

Note: This project is for educational and demonstration purposes only. It must not be used for clinical decision-making.

---

Jupyter Notebook

For a full walkthrough with code, outputs, and visualizations, see the Jupyter Notebook BreastCancerPredictAI.ipynb

Run the notebook online (no setup required):

---

Project Workflow

The pipeline follows these steps:

1. Data loading & cleaning

2. Exploratory Data Analysis (EDA)

   • Class distribution of malignant vs benign
   • Feature distributions and correlations

3. Preprocessing

   • Scaling
   • Train/test split with stratification

4. Model Training

   • Logistic Regression with GridSearchCV for hyperparameter tuning
   • Pipelines with ColumnTransformer, StandardScaler

5. Threshold Selection for Recall

   • Precision-recall analysis
   • Cross-validation to pick a threshold ensuring ~99% recall
   • Prioritizes recall to minimize false negatives in diagnosis context

6. Model Evaluation

   • Metrics: Accuracy, Precision, Recall, F1-score, ROC-AUC
   • Confusion matrix, precision-recall curves

7. Saving Model & Threshold

   • Best model saved with joblib
   • Optimal threshold stored in JSON for reproducibility