AI/ML coursework and student AI projects (machine learning, image emotion recognition, and misinformation diffusion analysis)
Developed and evaluated machine learning and deep learning pipelines using Python for multiple AI projects, including misinformation spread modeling and facial emotion recognition. Performed systematic dataset preprocessing and labeling-related transformations such as class-stratified cleaning and manual image preprocessing steps to prepare training inputs. Validated model performance using appropriate metrics (e.g., Macro F1 and AUC) and deployed results for real-world testing and feedback. • Implemented class-stratified cleaning and bot/seed-account detection in misinformation graph data preparation • Built custom SIR/SEIR simulator workflows to generate labeled outcomes across controlled experimental conditions • Created a manual preprocessing pipeline to standardize facial image inputs for 6-class emotion classification • Used class weighting/strategies for handling class imbalance during model training and evaluation