StratifyML — Hospital Analytics for Vulnerable Populations
- data science
- healthcare
- frequent itemset mining
- pandas
What it is
An analysis pipeline that studies hospital utilization across 13 vulnerable population groups (e.g. homeless, Medicaid, uninsured, elderly, rural).
It builds vulnerability flags, computes the Charlson Comorbidity Index from ICD-10 codes, then runs frequent-itemset mining (Apriori) over CPT/ICD code combinations and links those patterns to outcomes — length of stay, total charges, and mortality. Outputs publication-ready figures, LaTeX/CSV tables, and Excel reports highlighting health disparities.
Stack
Python · pandas · mlxtend (Apriori) · matplotlib/seaborn · networkx