Artificial Intelligence: Evaluating AI, optimizing AI

Clustering validation with ROC Curves

R Shiny platform for miRNA qPCR analysis in renal cell carcinoma: ΔΔCt normalization, geNorm stability, Wilcoxon testing, ROC-AUC, elastic net multi-miRNA classifier

This project builds a robust binary classification system to detect phishing websites using URL features like SSL state, web traffic, and domain patterns. Achieving 97% accuracy, 96.8% precision, and 0.9953 ROC-AUC, it ensures reliable, efficient detection with interpretable insights.

Comprehensive ML case study predicting cardiovascular disease using UCI dataset. Implemented Logistic Regression (87% accuracy) & KNN (89% AUC) with advanced feature engineering, correlation analysis & clinical insights for healthcare decision support systems.

Clinical trial patient screening pipeline built in Orange Data Mining. Predicts admission urgency (Elective, Urgent, Emergency) from 10,000 patient records using Neural Networks and Gradient Boosting. Includes ROC analysis, feature importance ranking, and a research paper on AI in healthcare knowledge management.

Deep learning for biomedical signals: HRV-based apnea detection and blood pressure estimation using hybrid CNN–LSTM–MLP models (GTS880)

Comparative analysis of Gaussian vs Gaussian Mixture Models for MNIST digit classification using MLE. Features 10-class ROC analysis, accuracy comparisons, and comprehensive model evaluation.

This repository contains many jupyter notebooks and output printouts chronicling my learning of basic Machine Learning concepts.

Machine learning project using the Wisconsin Diagnostic Breast Cancer dataset to classify tumors as malignant or benign. Includes data preprocessing, feature selection to reduce multicollinearity, and an interpretable decision tree model evaluated with stratified cross-validation, recall-focused tuning, and ROC analysis.

My DATAML200 - Pattern Recognition And Machine Learning course implementations.

Time to event analysis for neural data and motor units used in applied Brain machine interface (BMI)

Shiny App to help you generate adjusted ROC curves