papembengue0002-bit/HealthConnect-Experience-Lab-Week5-Data-Science

Week 5 Data Science deliverables for the AnalystLab Africa HealthConnect Experience Lab: patient-level no-show prediction using Logistic Regression, feature engineering, evaluation metrics, and actionable insights. README.md # HealthConnect Experience Lab - Week 5 Data Science ## Deliverables - `Week5_Data_Science_Baseline_Modelling.ipynb` - baseline modelling notebook - `Week5_Project_Summary.md` - required concise summary - `data/` - a separate copy of the provided model-ready source and derived features - `figures/` - five decision-supporting visualisations - `evaluation_metrics.json` - reproducible initial metrics - `src/week5_analysis.py` - dependency-free reproducible analysis implementation ## Model Logistic Regression predicts no-show risk using booking-time and prior-attendance information. The test split has no patient overlap with the training split. Initial ROC-AUC: **0.676**. This is a synthetic-data baseline, not a deployable clinical model.