This project is dedicated to the comparative analysis of medical visits (27,000+ records) using three different technology stacks. The main goal is to validate data and calculate key metrics of a medical facility's performance.

Built an interactive Healthcare Dashboard in Power BI to track patient trends, hospital performance, and key healthcare KPIs using DAX, Power Query, and data visualization.

My data analytics portfolio home

End-to-end healthcare analytics project using synthetic patient data. Performs exploratory analysis, feature engineering, and multi-class machine learning to predict medical test outcomes (Normal, Abnormal, Inconclusive) while generating interpretable insights for healthcare decision support.

SQL analysis of Alzheimer's biomarkers using ADNI data

Analysis of Hospital Admission Data Using SQL

The project aims to predict the likelihood of heart disease in individuals based on medical data using various machine learning algorithms. The project involves data preprocessing, feature selection, model training, and evaluation. The goal is to create a predictive model that assists healthcare professionals in diagnosing heart conditions

Exploratory Data Analysis on Diabetes dataset.

Exploratory data analysis and visualization of a heart disease dataset using Python, Pandas, Matplotlib, and Seaborn.

SQL-based healthcare analytics project with executive dashboards, hospital performance rankings, doctor analysis, revenue trends, and KPI metrics.

Interactive Power BI dashboard analysing hospital admissions, emergency vs planned admissions, wait times, and speciality performance using healthcare data.

Unsupervised clustering of NHANES biomarkers to test whether gestational diabetes history predicts distinct early cardiometabolic phenotypes in reproductive-age women

This project explores a synthetic healthcare dataset using Python and pandas. The data is loaded, cleaned, and analyzed through a full EDA pipeline, then visualized to uncover patterns across diagnosis, age, gender, and treatment.

Machine learning capstone using 253,680 CDC BRFSS records to evaluate kNN for diabetes classification in imbalanced health data. Tested scaling, SMOTE, Chi-Square feature selection, distance metrics, weighting strategies, and tree-based baselines; the optimized kNN achieved 0.88 ROC-AUC and 0.88 recall.

Healthcare data analytics project analysing hospital encounter data to evaluate clinical utilisation, length of stay, readmissions, payer exposure, and mortality trends.