Orta Doğu Teknik Üniversitesi Uygulamalı Matematik Enstitüsü Seminerleri
Controlled Markov Decision Processes with AVaR Criteria for Unbounded Cost
Kerem Uğurlu
University of Washington, Amerika Birleşik Devletleri
In this talk, we consider the control problem with the Average-Value-at-Risk (AVaR) criteria of the possibly unbounded L1-costs in infinite horizon on a Markov Decision Process (MDP). With a suitable state aggregation and by choosing a priori a global variable s heuristically, we show that there exist optimal policies for the infinite horizon problem for possibly unbounded costs.
