TD-regularized actor-critic methods

S Parisi, V Tangkaratt, J Peters, ME Khan - Machine Learning, 2019‏ - Springer
Actor-critic methods can achieve incredible performance on difficult reinforcement learning
problems, but they are also prone to instability. This is partly due to the interaction between …

Path consistency learning in tsallis entropy regularized mdps

Y Chow, O Nachum… - … conference on machine …, 2018‏ - proceedings.mlr.press
We study the sparse entropy-regularized reinforcement learning (ERL) problem in which the
entropy term is a special form of the Tsallis entropy. The optimal policy of this formulation is …

[PDF][PDF] Risk-Sensitive Reinforcement Learning with φ-Divergence-Risk

X Ni, L Lai - IEEE Transactions on Information …, 2024‏ - faculty.engineering.ucdavis.edu
Standard reinforcement learning (RL) algorithms primarily focus on minimizing the expected
sum of costs, which can be insufficient in contexts where risk sensitivity is crucial. This paper …

Divergence-augmented policy optimization

Q Wang, Y Li, J **ong, T Zhang - Advances in Neural …, 2019‏ - proceedings.neurips.cc
In deep reinforcement learning, policy optimization methods need to deal with issues such
as function approximation and the reuse of off-policy data. Standard policy gradient methods …

Co-evolution of predator-prey ecosystems by reinforcement learning agents

J Park, J Lee, T Kim, I Ahn, J Park - Entropy, 2021‏ - mdpi.com
The problem of finding adequate population models in ecology is important for
understanding essential aspects of their dynamic nature. Since analyzing and accurately …

f-Betas and portfolio optimization with f-divergence induced risk measures

R Ding - Quantitative Finance, 2023‏ - Taylor & Francis
In this paper, we build on using the class of f-divergence induced coherent risk measures for
portfolio optimization and derive its necessary optimality conditions formulated in CAPM …

Model-based reinforcement learning via stochastic hybrid models

H Abdulsamad, J Peters - IEEE Open Journal of Control …, 2023‏ - ieeexplore.ieee.org
Optimal control of general nonlinear systems is a central challenge in automation. Enabled
by powerful function approximators, data-driven approaches to control have recently …

[PDF][PDF] Path consistency learning in tsallis entropy regularized mdps

O Nachum, Y Chow… - arxiv preprint arxiv …, 2018‏ - proceedings.mlr.press
We study the sparse entropy-regularized reinforcement learning (ERL) problem in which the
entropy term is a special form of the Tsallis entropy. The optimal policy of this formulation is …

[PDF][PDF] Risk-Sensitive Reinforcement Learning with Coherent Risk Measures

X NI - 2025‏ - faculty.engineering.ucdavis.edu
Reinforcement Learning (RL) is a branch of machine learning that focuses on training
agents to make sequential decisions. By interacting with the environment, an RL agent …

[HTML][HTML] Statistical Machine Learning for Modeling and Control of Stochastic Structured Systems

H Abdulsamad - 2022‏ - tubiblio.ulb.tu-darmstadt.de
Machine learning and its various applications have driven innovation in robotics, synthetic
perception, and data analytics. The last decade especially has experienced an explosion in …