This repository is a case study, analysis and visualization of COVID-19 Pandemic spread along with prediction models.

Source code for predicting Blood Glucose Concentration

A causal convolutional neural network (CNN) model for predicting time series

Master's thesis work aims to develop a tool for the analysis and prediction of data from the MIMIC-III database, using sepsis as a case of study.

Production-grade ensemble framework combining XGBoost, PyTorch & Sklearn - 70%+ test coverage with Optuna optimization for time-series prediction

This repo focuses on latency-aware resource optimization for Kubernetes

Income prediction model based on U.S. Census dataset

Predicting Housing Market Conditions using PySpark and Machine Learning

A collection of deep learning solutions using PyTorch and TensorFlow covering Fashion-MNIST classification, FIFA match prediction, and Latin alphabet recognition.

Machine learning model for aviation accident risk prediction using ACAS and air traffic complexity data. Includes ONNX model export for deployment.

A repository which consists all of my machine model outputs done through different machine learning activities and assessments.

Tabular ML pipeline for TBI prediction models — loading, harmonization, feature engineering

Heart Disease Prediction using Logistic Regression, LDA, QDA, and KNN with feature engineering. LDA achieved best performance (82% accuracy, 94% recall). Focused on minimizing false negatives using ROC & Precision-Recall analysis for reliable medical classification.

This project aims to develop an AI-powered system for scheduling patient appointments based on medical history, symptoms, and doctor availability. It includes synthetic dataset generation, exploratory data analysis (EDA), and a machine learning model to predict consultation chains, optimizing healthcare workflow for efficient patient care.

TBI prediction model benchmarks — reproducible comparisons against IMPACT Core and CRASH Basic baselines

The Prediction Gap: 70% of Significant Meta-Analyses Have Null-Spanning Prediction Intervals

Open research notebooks for World Cup 2026 football data analysis, baseline models, and reproducible evaluation.

A user-friendly web application for performing data analysis with machine learning techniques, built with Streamlit

This project was built not just for a grade, but with the hope that it could help raise awareness about diabetes risk and inspire others to learn, share, and build technology that cares.

Optimizing bets on political elections