I am a passionate Researcher specializing in Trustworthy AI and Healthcare AI. My focus is on Data Science and AI Research.

Evolutionary Computing

• finchGE A Modular Grammatical Evolution Library : FinchGE is a modern, modular, and user-friendly Python library for Grammatical Evolution (GE)- a powerful evolutionary algorithm that uses formal grammars to evolve programs, expressions, and solutions.

• finchge-examples Example projects for demonstration of FinchGE API Usage.

AI/ML in Healthcare

• wearable-health-llmops A health prediction system that uses wearable sensor features to estimate readiness with either classical ML models or local LLMs through a containerized API and interactive UI.

• cohort-selection-criteria-extractor A rule-based approach for extracting structured cohort selection information from PubMed Central full-text NXML articles, achieving the score 0.74096 on the Kaggle task: CohortX Task 1: Extracting Cohort Selection Criteria from Free-Texts | Kaggle.

• MicroRNA-Signatures-for-Neurodegenerative-Diseases – Analysis for identifying microRNA signatures associated with various neurodegenerative diseases using publicly available gene expression datasets.

• MicroRNA-Signatures-for-Neurodegenerative-Diseases – Analysis for identifying microRNA signatures associated with various neurodegenerative diseases using publicly available gene expression datasets.

• microRNA-dementia-risk-prediction – Machine learning system for predicting dementia risk using serum miRNA biomarkers.

• ML-for-Cardiovascular-Disease-Diagnosis – Comparison of performance of various ML models on binary classification of cardiovascular disease (CVD) using four distinct public datasets. (Aiflow)

• spirometry-risk-prediction-nhanes - Implementation of segmented regression to model spirometry measures with Exploratoory Data Analysis EDA, feature engineering for clinical data interpretation.

• multiomics-survival-prediction - Predicting survival in head and neck cancer patients using clinical data, genetic information, and CT scan analysis. Combines multiple data types to build better prognostic models.

• SmokingStatusPrediction4COPD - Supplimentary Information for the paper: Development of a Deep Learning Model to Predict Smoking Status in Patients with Chronic Obstructive Pulmonary Disease: A Secondary Analysis of Cross-sectional National Survey

Data Engineering & Pipelines

• airflow-demo – heart_risk_pred_pipeline using Airflow
• pdfchat-rag - a modular, containerized application that enables semantic search and chat over PDF documents using RAG. Built with Flask, Celery, Ollama, ChromaDB

Misc

• ETRI-Human-Understanding-AI-Paper-Competition - About EDA, Preprocessign and Model Training code for The 2nd ETRI Human Understanding Artificial Intelligence Paper Competition.

• DSS_challenges - Projects for Data Science Syndicate Challenges

• onedr - A lightweight python library for drone control using MAVLINK.