Tag Archives: learning. Markov

Random  design #19

A random walk sounds like fun. Who hasn’t tried it before?

This is the result of 250 random walks around a 32×31 grid where walks were limited to 3050 steps. And guess what happened? Art having no boundaries – I thanked Jim Kan for setting up the game and walked with the board in a higher (mathematical) dimension!

Jim’s original idea is based on SARSA, an algorithm for learning a Markov decision process. For the more adventurous, I encourage you to visit his page, it is quite impressive – and very colorful too.