Risk-Sensitive Reinforcement Learning: Near-Optimal Risk-Sample Tradeoff in Regret

by   Yingjie Fei, et al.

We study risk-sensitive reinforcement learning in episodic Markov decision processes with unknown transition kernels, where the goal is to optimize the total reward under the risk measure of exponential utility. We propose two provably efficient model-free algorithms, Risk-Sensitive Value Iteration (RSVI) and Risk-Sensitive Q-learning (RSQ). These algorithms implement a form of risk-sensitive optimism in the face of uncertainty, which adapts to both risk-seeking and risk-averse modes of exploration. We prove that RSVI attains an Õ(λ(|β| H^2) ·√(H^3 S^2AT)) regret, while RSQ attains an Õ(λ(|β| H^2) ·√(H^4 SAT)) regret, where λ(u) = (e^3u-1)/u for u>0. In the above, β is the risk parameter of the exponential utility function, S the number of states, A the number of actions, T the total number of timesteps, and H the episode length. On the flip side, we establish a regret lower bound showing that the exponential dependence on |β| and H is unavoidable for any algorithm with an Õ(√(T)) regret (even when the risk objective is on the same scale as the original reward), thus certifying the near-optimality of the proposed algorithms. Our results demonstrate that incorporating risk awareness into reinforcement learning necessitates an exponential cost in |β| and H, which quantifies the fundamental tradeoff between risk sensitivity (related to aleatoric uncertainty) and sample efficiency (related to epistemic uncertainty). To the best of our knowledge, this is the first regret analysis of risk-sensitive reinforcement learning with the exponential utility.


page 1

page 2

page 3

page 4


Bridging Distributional and Risk-sensitive Reinforcement Learning with Provable Regret Bounds

We study the regret guarantee for risk-sensitive reinforcement learning ...

Regret Bounds for Risk-Sensitive Reinforcement Learning

In safety-critical applications of reinforcement learning such as health...

Variational Bayesian Reinforcement Learning with Regret Bounds

We consider the exploration-exploitation trade-off in reinforcement lear...

Risk-Sensitive Reinforcement Learning: a Martingale Approach to Reward Uncertainty

We introduce a novel framework to account for sensitivity to rewards unc...

Efficient exploration via epistemic-risk-seeking policy optimization

Exploration remains a key challenge in deep reinforcement learning (RL)....

Exponential Bellman Equation and Improved Regret Bounds for Risk-Sensitive Reinforcement Learning

We study risk-sensitive reinforcement learning (RL) based on the entropi...

Tiered Reinforcement Learning: Pessimism in the Face of Uncertainty and Constant Regret

We propose a new learning framework that captures the tiered structure o...

Please sign up or login with your details

Forgot password? Click here to reset