Faithkanyuki/Moringa_CapStone_Project

Optimizing Healthcare Resource Allocation and Patient Outcomes for Diabetic Patients in Kenya. 📑 Table of Contents Introduction Project Overview Phase 1: Business Understanding Phase 2: Data Understanding Phase 3: Data Preparation & Feature Engineering Phase 4: Exploratory Data Analysis (EDA) Readmission Rate by Categorical Variables Distributions of Numerical Variables by Readmission Status Correlation Matrix of Numerical Features Readmission Rate by Diagnosis Groups Readmission Rate by Time in Hospital Phase 5: Modeling Logistic Regression Random Forest XGBoost Phase 6: Model Evaluation & Comparison Phase 7: Final Model Selection Phase 8: Business Impact & Recommendations Deployment Conclusion Recommendations Tools & Technologies Authors ### Diabetes Readmission Risk Prediction for Kenya's Healthcare System _Knowing early. Acting wisely. Live longer_ kenya-hospital-streamlit.on… # Introduction Diabetes poses a growing burden on Kenya’s healthcare system, where limited resources must serve an increasing number of patients. Inefficient allocation of healthcare resources often leads to preventable complications, hospital readmissions, and increased costs. This project explores the use of machine learning techniques to analyze patient data and predict healthcare needs, enabling data-driven decision-making. By identifying high-risk diabetic patients early, the study aims to support targeted interventions, optimize resource utilization, and improve patient outcomes within Kenya’s constrained healthcare environment. ## Project Overview Diabetes-related hospital readmissions place significant strain on Kenya’s healthcare system by increasing costs, overcrowding hospitals, and stretching already limited healthcare resources. This project applies **machine learning techniques** to predict **30-day hospital readmission risk** among diabetic patients. The goal is to enable **early identification of high-risk patients**, support **targeted interventions**, and improve **healthcare resource allocation**. The project follows the **CR …