The algorithm works by reducing the tree-structured game to a linear bandits problem and then applying some variant of EXP-IX (similar to Jin et al., ...
確定! 回上一頁