Parallel nonstationary direct policy search for risk-averse stochastic optimization

Somayeh Moazeni, Warren Buckler Powell, Boris Defourny, Belgacem Bouzaiene-Ayari

Research output: Contribution to journalArticlepeer-review

6 Scopus citations


This paper presents an algorithmic strategy to nonstationary policy search for finite-horizon, discrete-time Markovian decision problems with large state spaces, constrained action sets, and a risk-sensitive optimality criterion. The methodology relies on modeling time-variant policy parameters by a nonparametric response surface model for an indirect parametrized policy motivated by Bellman's equation. The policy structure is heuristic when the optimization of the risk-sensitive criterion does not admit a dynamic programming reformulation. Through the interpolating approximation, the level of nonstationarity of the policy, and consequently, the size of the resulting search problem can be adjusted. The computational tractability and the generality of the approach follow from a nested parallel implementation of derivative-free optimization in conjunction with Monte Carlo simulation. We demonstrate the efficiency of the approach on an optimal energy storage charging problem, and illustrate the effect of the risk functional on the improvement achieved by allowing a higher complexity in time variation for the policy.

Original languageEnglish (US)
Pages (from-to)332-349
Number of pages18
JournalINFORMS Journal on Computing
Issue number2
StatePublished - Mar 1 2017

All Science Journal Classification (ASJC) codes

  • Software
  • Information Systems
  • Computer Science Applications
  • Management Science and Operations Research


  • Derivative-free optimization
  • Direct policy search
  • Dynamic optimization
  • Energy storage
  • Learning
  • Parallel optimization
  • Risk-averse stochastic optimization


Dive into the research topics of 'Parallel nonstationary direct policy search for risk-averse stochastic optimization'. Together they form a unique fingerprint.

Cite this