A crucial aspect of intelligence is the ability to reason about what actions to take to reach a specific goal. Planning is the area of Artificial Intelligence that studies how to find such a sequence of actions. Planning problems, however, are often difficult or even infeasible to solve.

Several approaches exist that use domain knowledge to make practical planning possible. The goal of this project is to explore a framework for representing knowledge that unifies several different planning strategies under a single representation, using probabilistic statements. We will develop a planner that uses these statements to trade-off solution quality and run-time, and a learning algorithm that uses data from previous problems to find structure.

I have been working with the LIS group for the past 3 years, doing research on Artificial Intelligence, Robotics, and Machine Learning. This project will be a direct continuation of my 6.UAP project, which I worked on at this lab.