methodology-2

Multi-Objective Reinforcement Learning

3260 papers • 126 benchmarks • 313 datasets

This task has no description! Would you like to contribute one?

(Image credit: Papersgraph)

Benchmarks

These leaderboards are used to track progress in multi-objective-reinforcement-learning-5

Trend

Dataset

Best Model

Actions

No benchmarks available.

Libraries

i

Use these libraries to find multi-objective-reinforcement-learning-5 models and implementations

lucasalegre/morl-baselines

8 papers 220

Datasets

No datasets available.

Subtasks

No subtasks available.

Most implemented papers

Optimization of Molecules via Deep Reinforcement Learning

Zhenpeng Zhou, S. Kearnes, Li Li, R. Zare, Patrick F. Riley•Thu Oct 18 2018

Inspired by problems faced during medicinal chemistry lead optimization, the MolDQN model is extended with multi-objective reinforcement learning, which maximizes drug-likeness while maintaining similarity to the original molecule.

618

Content

lucasalegre/mo-gym

3 papers 238

0

Paper Graph

Multi-Objective Deep Reinforcement Learning

Shimon Whiteson, Yannis Assael, D. Roijers, Hossam Mossalam•Sat Oct 08 2016

This work proposes Deep Optimistic Linear Support Learning (DOL) to solve high-dimensional multi-objective decision problems where the relative importances of the objectives are not known a priori.

172 0

Paper Graph

Anchor-Changing Regularized Natural Policy Gradient for Multi-Objective Reinforcement Learning

Ruida Zhou, Tao-Wen Liu, D. Kalathil, P. Kumar, Chao Tian•Thu Jun 09 2022

An Anchor-changing Regularized Natural Policy Gradient (ARNPG) framework, which can systematically incorporate ideas from well-performing first-order methods into the design of policy optimization algorithms for multi-objective MDP problems, is proposed.

16 0

Paper Graph

A Generalized Algorithm for Multi-Objective Reinforcement Learning and Policy Adaptation

Xingyuan Sun, Runzhe Yang, Karthik Narasimhan•Tue Aug 20 2019

A generalized version of the Bellman equation is proposed to learn a single parametric representation for optimal policies over the space of all possible preferences in MORL, with the goal of enabling few-shot adaptation to new tasks.

315 0

Paper Graph

Dynamic Weights in Multi-Objective Deep Reinforcement Learning

Denis Steckelmacher, A. Nowé, D. Roijers, Axel Abels, T. Lenaerts•Wed Sep 19 2018

This work proposes a multi-objective Q-network whose outputs are conditioned on the relative importance of objectives, and introduces Diverse Experience Replay (DER) to counter the inherent non-stationarity of the dynamic weights setting.

191 0

Paper Graph

Sample-Efficient Multi-Objective Learning via Generalized Policy Improvement Prioritization

L. N. Alegre, A. Bazzan, Bruno C. da Silva, D. Roijers, Ann Now'e•Tue Jan 17 2023

A novel algorithm that uses Generalized Policy Improvement (GPI) to define principled, formally-derived prioritization schemes that improve sample-efficient learning and empirically shows that the method outperforms state-of-the-art MORL algorithms in challenging multi-objective tasks, both with discrete and continuous state and action spaces.

48 0

Paper Graph

MO-Gym: A Library of Multi-Objective Reinforcement Learning Environments

A. Nowé, El-Ghazali Talbi, L. N. Alegre, Florian Felten, Grégoire Danoy, A. Bazzan, Bruno C. da Silva•Sat Dec 31 2022

MO-Gym is introduced, an extensible library containing a diverse set of multi-objective reinforcement learning environments that extends the widely-used OpenAI Gym API, allowing the reuse of algorithms and features that are well-established in the reinforcement learning community.

30 0

Paper Graph

Predicting optimal value functions by interpolating reward functions in scalarized multi-objective reinforcement learning

A. Kusari, J. How•Tue Sep 10 2019

It is proved that the value function transforms smoothly given a transformation of weights of the reward function (and thus a smooth interpolation in the policy space) and that the interpolation can provide robust values for sample states and actions in both discrete and continuous domain problems.

4 0

Paper Graph

A Distributional View on Multi-Objective Policy Optimization

H. F. Song, N. Heess, Martin A. Riedmiller, R. Hadsell, A. Abdolmaleki, M. Neunert, Sandy H. Huang, Leonard Hasenclever, Martina Zambelli, M. Martins•Thu May 14 2020

This paper proposes a novel algorithm for multi-objective reinforcement learning that enables setting desired preferences for objectives in a scale-invariant way, and uses supervised learning to fit a parametric policy to a combination of these distributions.

89 0

Paper Graph

Adding a benchmark result helps the community track progress.