@shanive, I've implemented a Hoeffding-estimate based agent (HCT) which chooses a switch with maximum value-of-information, I'll let you read the paper draft later.
Please run the tournamet for HCT, UCT, GCT, let me know the results, only for winloss.
@shanive, I've implemented a Hoeffding-estimate based agent (HCT) which chooses a switch with maximum value-of-information, I'll let you read the paper draft later.
Please run the tournamet for HCT, UCT, GCT, let me know the results, only for winloss.