Machine Learning
Heart Failure Survival Prediction
Clinical ML system predicting patient survival using GMM-based synthetic augmentation and ensemble classifiers.
The Problem
Context & Challenge
Heart failure has high mortality, but clinical datasets are small. The challenge was building an accurate survival prediction model despite limited training data (299 patients).
The Approach
Architecture & Implementation
Tested 21 model configurations across Logistic Regression, KNN, Random Forest, Gradient Boosting, and Neural Networks. Addressed small-sample limitations with GMM-based synthetic data augmentation (5,299 generated samples) and threshold tuning to minimize missed high-risk cases.
The Results
Impact & Metrics
Achieved 85% test accuracy with Gradient Boosting on augmented data. Synthetic augmentation reduced test loss by 42% compared to original dataset training.
Key Result
85% test accuracy with 42% loss reduction via synthetic data augmentation
Technologies & Methods
PythonTensorFlow/Kerasscikit-learnGradient BoostingRandom ForestGMMThreshold Tuning