Detecting vehicles on the road

Neural-net based optimization for bucket operation

We design novel algorithms to solve very large scale data mining problems from rich and diverse information sources. Our techniques for active learning and experiment design allow us to maximize the benefit of expert labeling while minimizing its time and cost. We apply this research very broadly. Our machine learning applications include classifying terrain from aerial multi-spectral imagery, detecting obstacles for robotic applications in challenging terrain, modeling and predicting the motion of skid-steered vehicles, identifying cars and trucks on the highway, and inspecting and grading strawberry plants.

We focus particularly on machine learning for problems of decision mining, reinforcement learning, and learning control. In these applications, the learning algorithm's decision has an immediate effect on the environment it interacts with. Our algorithms have been applied to the optimal control of robotic manipulators, learning-based obstacle avoidance, route-planning and re-planning, and automated drug discovery.