Bayesian Belief Network Simulation - Alien Invasion!

This simulation is part of my series on interactive AI education. I use this simulation game in my class to help demonstrate the algorithms we are learning. The code (not the art assets, though) is available on my GitHub (username MLivanos).

Aliens are coming to Davis, California, and we need your help to better understand what, if any, the threat is. In this interactive demonstration, we will build a Bayesian Belief Network to model the city of Davis, California with the potential for alien visitation. We will use the model to reason about when the aliens will come and what the warning signs will be to best prepare for these encounters.

This simulation explores Bayesian Belief Networks, an efficient model of joint probability distributions. Unlike problems in a textbook, this network is too large for exact with 10 nodes and 17 connections. Instead, we rely on sampling methods - rejection sampling and likelihood weighting - to help us estimate the probability of visitation with the information we know.