Building generative AI applications for various use cases

Advisor: Prof. Chenhui Shao- Conducted research focused on advanced machine learning methods for smart manufacturing- Demonstrated the first applications of federated learning in additive manufacturing and machine fault diagnosis- Formulated a novel clustering algorithm to tackle on-device statistical heterogeneity in federated learning- Developed an end-to-end response surface methodology for multi-task learning with input-dependent noise- Formulated a novel framework to… Show more Advisor: Prof. Chenhui Shao- Conducted research focused on advanced machine learning methods for smart manufacturing- Demonstrated the first applications of federated learning in additive manufacturing and machine fault diagnosis- Formulated a novel clustering algorithm to tackle on-device statistical heterogeneity in federated learning- Developed an end-to-end response surface methodology for multi-task learning with input-dependent noise- Formulated a novel framework to combine active learning and multi-task learning for Gaussian processes- Enabled highly data-efficient, flexible, and accurate modeling in crucial manufacturing applications like defect detection from images, fault diagnosis from sensor signals, part qualification after 3D printing, and surface shape prediction Show less

- Formulated and deployed ML models for wafer metrology using sensor data- Improved the performance of existing ML models by over 80% in Seagate factories

- Automated the detection of Fluorine deposition depth from images- Performed statistical analysis of machine age data for remaining useful life prediction- Programmed a physics-informed heat transfer solver for deposition tools

- Final year project with the Computer-Aided Engineering team- Built CFD models, wrote Python scripts to analyze and infer experimental data