A repo with code for the three programming elements of problem sets for Harvard's CS182 course on Artificial Intelligence.
-
Pset0 is an implementation of the classic travelling salesman problem to figure out the time to travel between different metros stations in Boston. Practically, it uses a "breadth first search" to find the minimum distance between two vertices in an adjacency matrix graph. (Grade: 4/5) 💼
-
Pset1 implements a Markov decision process (MDP), value Iteration and policy Iteration, and belief state updates for a partially observable Markov decision process (POMDP). (Grade: 4/7) 🌝
-
Pset3 solves a well-known robotic control problem – landing a lunar agent on the surface of the moon – using reinforcement learning (RL). Practically, it uses the Lunar Lander environment in OpenAI Gym using Q-learning, a model-free RL algorithm. (Grade: 5/6) 🚀