Hacker News Reader

Nothing so advanced, I'm afraid. The Monte Carlo system just creates a thousand points (2d position and velocity), and simulates them at 1 generation per second, culling points that are unlikely and reproducing points that are likely based on observations (clicking chunks and hearing back "loaded" or "unloaded").

View on HN

That sounds a lot like sequential monte carlo or a particle filter, which is a type of bayesian filter for inference. Really impressive what you did.

The write up actually does reference a paper on particle filters :o

Yeah, https://users.isy.liu.se/rt/schon/Publications/HolSG2006.pdf was great to refer to, just the right level of complexity for me.

Reply on news.ycombinator.com