Automated receipt categorization using a swarm intelligence multi-agent system. Features annotation-driven extraction, adaptive consensus, and robust expense classification for real-world document processing.

This Repository deals with Credit Risk evaluation and prediction of default

Deep Learning-based classification of diabetic kidney disease from histopathology images using CNN models and medical image analysis techniques.

A web application for predicting the species of Iris flowers using a machine learning model trained with the Iris dataset, with FastAPI, a modern web framework for building APIs.

Ensemble gradient-boosting models for fatigue prediction with calibrated prediction intervals and data-split comparison

Trains a custom ensemble model from scratch on the top 10 crypto coins by market cap and ranks them by predicted move + confidence over the next 1-2 hours. No pretrained weights, no cached model - retrains every run.

End-to-end Data Science & ML projects covering Healthcare, Retail, Automotive & Finance — EDA, feature engineering, regression, classification, and statistical analysis | Python · Scikit-learn · Pandas