We develop a new method for stochastic optimization using Bayesian statistics approach. More precisely, we optimize parameters of chess engines as those data are available to us, but the method should be applicable to all situations where we want to optimize some gain/loss function which has no analytical form and thus cannot be measured directly, but only by comparison of two parameter sets. We also experimentally compare the new method with the famous SPSA method.
| Bayesian statistics approach to chess engines optimization | 549.9KB |