We're blogging machines!
Subscribe to feed
‹ Interesting Fact about Student T distributions • Standard Deviation of Sample Median ›
July 26, 2012 in Ensemble Learning, Multi-Armed Bandit Problem, Optimization by hundalhh | Permalink
I’m rather fascinated by the HOO algorithm. It seems to be a combination of simulated annealing, branch and bound, simplex method, and upper confidence trees. I’m going to code it up an try it on a few problems.
Powered by WordPress and Tarski