This repository contains the source code and resources for my research project on early stroke detection using machine learning techniques. The project leverages feature selection methods such as Recursive Feature Elimination (RFE) and Permutation Feature Importance (PFI) to build interpretable, reliable models that can assist in medical diagnosis.
To read more about the analysis: The Use of Machine Learning in Stroke Detection: Performance, Medicinal Knowledge, and Future Prospects (2024 Dec). [https://doi.org/10.70121/001c.127432]
- ✅ Preprocessing pipeline for medical datasets (clinical + lifestyle factors)
- ✅ Feature selection using RFE and PFI
- ✅ Multiple ML models implemented (Logistic Regression, Random Forest, etc.)
- ✅ Evaluation with accuracy, precision, recall, F1-score, and other metrics
- ✅ Research-backed methodology validated in a peer-reviewed publication