Breast Cancer Prediction
Supervised learning models to predict tumor malignancy with high precision from medical data.
Overview
Built a predictive model to classify benign vs. malignant tumors using breast cancer diagnostic datasets. The project combined careful data preprocessing with supervised ML algorithms to support medical decision-making.
Approach
- Data Cleaning & EDA: Explored tumor features (clump thickness, mitosis, etc.).
- Feature Scaling & Selection to improve model generalization.
- Modeling: Logistic Regression, Random Forests, and SVMs.
- Hyperparameter Tuning for optimization.
- Deployment: Prototyped API deployment using Flask/FastAPI for practical use cases.
Results
- Achieved high precision and recall across multiple ML models.
- Produced a deployable prototype for clinical decision support.
Skills demonstrated: classification modeling, healthcare data analysis, deployment prototyping, precision/recall optimization.