My research is in solving hard problems by adding learning components to existing approaches. For me, this usually means adding artificial neural networks or other numerical methods to model the existing solution or to discover new solutions via reinforcement learning. Also see the summary of the artificial intelligence program at Colorado State University.
Active Projects
- Reinforcement Learning and Control, (also see Robust Reinforcement Learning Control)
- EEG Pattern Classification, for Brain-Machine Interfaces
Past Projects
- PAPA a program for predicting the prion propensity of a protein. Javascript version translated from Python version by Asa Ben-Hur
- Modeling and Control of Algae Populations for Biodiesel Production
- Atmosphere Model Inversion with Neural Networks
- Computer Graphics (segmentation, NURBS, radiosity)
- Adaptive Training of Pilots
- Neural Modeling, as part of the Molecular, Cellular, and Integrative Neurosciences Program at CSU
- Software and Hardware Testing
- Computer Vision
- Parallel Implementations