Selected from 3,000+ applicants for a competitive fellowship focused on applied machine learning. Building production-grade ML systems and collaborating with industry partners.
Robot Learning, Perception, and Planning
Building ML models to distinguish stable and unstable regimes in high-dimensional dynamical systems. Reduced labeled data requirements by ~60% through boundary-region learning. Achieved 97% classification accuracy while preserving reliability near regime transition points.
Machine Learning for Genomic Analysis
Engineered ML pipeline analyzing 640M+ DNA bases to predict fungicide resistance in plant disease pathogens. Training transformer-based sequence models to learn resistance patterns from large-scale fungal genome data. Achieved ~73% prediction accuracy generalizing reliably across held-out genomic samples.