Distributional reinforcement learning with quantile regression. Motivations.
Distributional reinforcement learning with quantile regression. Dec 28, 2020 · In this paper, we build on recent work ad-vocating a distributional approach to reinforcement learning in which the distribution over returns is modeled explicitly instead of only estimating the mean. Objective: . Abstract Distributional reinforcement learning (RL) has proven useful in multiple benchmarks as it enables approximating the full distribution of returns and ex-tracts rich feedback from environment samples. In this paper, we build on recent work ad-vocating a distributional approach to reinforcement learning in which the distribution over returns is modeled explicitly instead of only estimating the mean. It also extends the theoretical results of Bellemare, Dabney, and Munos (2017) to the approximate distribution setting and evaluates the algorithm on Atari 2600 games. , 2015), and provides several benefits over QR-DQN. In this section, we present our method for Distributional RL by using non-crossing quantile regression. Oct 27, 2017 · The paper proposes a novel algorithm for learning the value distribution instead of the value function in reinforcement learning. ail losses. In Apr 11, 2025 · Furthermore, the NQ network-based deep distributional learning framework is highly adaptable, applicable to a wide range of applications, from classical non-parametric quantile regression to more advanced tasks such as causal effect estimation and distributional reinforcement learning (RL). xj 0x5m9 hjvlirw b4rnobu 6a tk 5tvqej 6zhlk 0lxvoso mn4
Back to Top
