What is intelligence? Is it possible for a computer to possess intelligence or be intelligence? How do we measure this quality about a non-human, or inanimate object? In this class, we will explore methods to solve problems that seemingly require intelligence.

The first portion of the class focuses on utilizing search techniques for solving problems. Approaches include uninformed search (BFS, DFS), informed heuristical search (A*, IDA), local search (hill climbers, monte carlo and simulated annealing search) and constraint satisfaction. The course then investigates developing agents for adversarial games, using MiniMax game tees, \(\alpha-\beta\) pruning, expectimax, markov decision processes (MDPs) and reinforcement learning (Q-Learning).

Knowledge bases (KB), which are sets of facts, are investigated to permit computations to infer new facts from existing ones through algorithmic entailment. This includes investigating the power of propositional and first-order logic (FOL) and the inference algorithms associated with these logics.

The last section of the course will focus on utilizing a probabilistic representation of the world, so that we can develop more advanced intelligent “agents”. These techniques includes Bayes nets and hidden Markov models (HMMs).