Logo Lanfrica

natzhhh/Interpretable-diabetes-prediction

Domaine:

healthcare

Type de record:

software
Créateur:
nat
Hôte:
This repository contains the source code for a Clinical Decision Support System (CDSS) developed for a graduate MSc thesis of Computer Science in Haramaya University, Ethiopia. # Interpretable AI for Diabetes Prediction (on Iraqi Population Diabetes Dataset) This repository contains the source code for a Clinical Decision Support System (CDSS) developed for a graduate thesis in Haramaya University, Ethiopia. ### Core Features: - **Global Explanations:** SHAP-based feature importance. - **Interventional Analysis:** Multi-directional counterfactuals (Recovery and Risk paths) using DiCE. - **Trust Framework:** Aligned with Asan et al. (2020) Trust in Healthcare AI. ### To Run Locally: 1. pip install -r requirements.txt 2. streamlit run app.py