Timezone: »
In many sequential decision-making problems we may want to manage risk by minimizing some measure of variability in costs in addition to minimizing a standard criterion. Conditional value-at-risk (CVaR) is a relatively new risk measure that addresses some of the shortcomings of the well-known variance-related risk measures, and because of its computational efficiencies has gained popularity in finance and operations research. In this paper, we consider the mean-CVaR optimization problem in MDPs. We first derive a formula for computing the gradient of this risk-sensitive objective function. We then devise policy gradient and actor-critic algorithms that each uses a specific method to estimate this gradient and updates the policy parameters in the descent direction. We establish the convergence of our algorithms to locally risk-sensitive optimal policies. Finally, we demonstrate the usefulness of our algorithms in an optimal stopping problem.
Author Information
Yinlam Chow (Google)
Mohammad Ghavamzadeh (Facebook AI Research)
More from the Same Authors
-
2019 Workshop: Safety and Robustness in Decision-making »
Mohammad Ghavamzadeh · Shie Mannor · Yisong Yue · Marek Petrik · Yinlam Chow -
2019 Poster: Tight Regret Bounds for Model-Based Reinforcement Learning with Greedy Policies »
Yonathan Efroni · Nadav Merlis · Mohammad Ghavamzadeh · Shie Mannor -
2019 Spotlight: Tight Regret Bounds for Model-Based Reinforcement Learning with Greedy Policies »
Yonathan Efroni · Nadav Merlis · Mohammad Ghavamzadeh · Shie Mannor -
2018 Poster: A Lyapunov-based Approach to Safe Reinforcement Learning »
Yinlam Chow · Ofir Nachum · Edgar Duenez-Guzman · Mohammad Ghavamzadeh -
2018 Poster: A Block Coordinate Ascent Algorithm for Mean-Variance Optimization »
Tengyang Xie · Bo Liu · Yangyang Xu · Mohammad Ghavamzadeh · Yinlam Chow · Daoming Lyu · Daesub Yoon -
2017 Poster: Conservative Contextual Linear Bandits »
Abbas Kazerouni · Mohammad Ghavamzadeh · Yasin Abbasi · Benjamin Van Roy -
2016 Poster: Safe Policy Improvement by Minimizing Robust Baseline Regret »
Mohammad Ghavamzadeh · Marek Petrik · Yinlam Chow -
2015 Workshop: Machine Learning for (e-)Commerce »
Esteban Arcaute · Mohammad Ghavamzadeh · Shie Mannor · Georgios Theocharous -
2015 Poster: Risk-Sensitive and Robust Decision-Making: a CVaR Optimization Approach »
Yinlam Chow · Aviv Tamar · Shie Mannor · Marco Pavone -
2015 Poster: Policy Gradient for Coherent Risk Measures »
Aviv Tamar · Yinlam Chow · Mohammad Ghavamzadeh · Shie Mannor -
2014 Workshop: Large-scale reinforcement learning and Markov decision problems »
Benjamin Van Roy · Mohammad Ghavamzadeh · Peter Bartlett · Yasin Abbasi Yadkori · Ambuj Tewari -
2013 Poster: Actor-Critic Algorithms for Risk-Sensitive MDPs »
Prashanth L.A. · Mohammad Ghavamzadeh -
2013 Poster: Approximate Dynamic Programming Finally Performs Well in the Game of Tetris »
Victor Gabillon · Mohammad Ghavamzadeh · Bruno Scherrer -
2013 Oral: Actor-Critic Algorithms for Risk-Sensitive MDPs »
Prashanth L.A. · Mohammad Ghavamzadeh -
2012 Poster: Best Arm Identification: A Unified Approach to Fixed Budget and Fixed Confidence »
Victor Gabillon · Mohammad Ghavamzadeh · Alessandro Lazaric -
2011 Poster: Multi-Bandit Best Arm Identification »
Victor Gabillon · Mohammad Ghavamzadeh · Alessandro Lazaric · Sebastien Bubeck -
2011 Poster: Speedy Q-Learning »
Mohammad Gheshlaghi Azar · Remi Munos · Mohammad Ghavamzadeh · Hilbert J Kappen -
2010 Spotlight: LSTD with Random Projections »
Mohammad Ghavamzadeh · Alessandro Lazaric · Odalric-Ambrym Maillard · Remi Munos -
2010 Poster: LSTD with Random Projections »
Mohammad Ghavamzadeh · Alessandro Lazaric · Odalric-Ambrym Maillard · Remi Munos -
2008 Workshop: Model Uncertainty and Risk in Reinforcement Learning »
Yaakov Engel · Mohammad Ghavamzadeh · Shie Mannor · Pascal Poupart -
2008 Poster: Regularized Policy Iteration »
Amir-massoud Farahmand · Mohammad Ghavamzadeh · Csaba Szepesvari · Shie Mannor -
2007 Spotlight: Incremental Natural Actor-Critic Algorithms »
Shalabh Bhatnagar · Richard Sutton · Mohammad Ghavamzadeh · Mark P Lee -
2007 Poster: Incremental Natural Actor-Critic Algorithms »
Shalabh Bhatnagar · Richard Sutton · Mohammad Ghavamzadeh · Mark P Lee -
2006 Poster: Bayesian Policy Gradient Algorithms »
Mohammad Ghavamzadeh · Yaakov Engel -
2006 Spotlight: Bayesian Policy Gradient Algorithms »
Mohammad Ghavamzadeh · Yaakov Engel