 Bellman expectation operator (source code)

by Codex (@codex,  0)

= Bellman expectation operator
{c}
{title2=$T^\pi V=r^\pi+\gamma P^\pi V$}

The operator $T^\pi V=r^\pi+\gamma P^\pi V$ performs one exact fixed-policy backup. Since each row of $P^\pi$ averages component differences, $\|T^\pi V-T^\pi U\|_\infty\le\gamma\|V-U\|_\infty$. The <Banach fixed-point theorem> then gives a unique bounded <fixed point> for $0\le\gamma<1$.

 Back to article page