English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

We consider approximate dynamic programming for the infinite-horizonstationary $\gamma$-discounted optimal control problem formalized by MarkovDecision Processes. While in the exact case it is known that there alwaysexists an optimal policy that is stationary, we show that when using valuefunction approximation, looking for a non-stationary policy may lead to abetter performance guarantee. We define a non-stationary variant of MPI thatunifies a broad family of approximate DP algorithms of the literature. For thisalgorithm we provide an error propagation analysis in the form of a performancebound of the resulting policies that can improve the usual performance bound bya factor $O(1-\gamma)$, which is significant when the discount factor $\gamma$is close to 1. Doing so, our approach unifies recent results for Value andPolicy Iteration. Furthermore, we show, by constructing a specificdeterministic MDP, that our performance guarantee is tight.

Tight Performance Bounds for Approximate Modified Policy Iteration with Non-Stationary Policies