Efficient Multi-objective Reinforcement Learning via Multiple-gradient Descent with Iteratively Discovered Weight-Vector Sets
From MaRDI portal
Publication:5145843
DOI10.1613/jair.1.12270zbMath1497.68414MaRDI QIDQ5145843
Publication date: 22 January 2021
Published in: Journal of Artificial Intelligence Research (Search for Journal in Brave)
Full work available at URL: https://doi.org/10.1613/jair.1.12270
90C29: Multi-objective and goal programming
68T05: Learning and adaptive systems in artificial intelligence
90C40: Markov and semi-Markov decision processes
Uses Software
Cites Work
- Unnamed Item
- Unnamed Item
- Unnamed Item
- Multiple-gradient descent algorithm (MGDA) for multiobjective optimization
- Online actor-critic algorithm to solve the continuous-time infinite horizon optimal control problem
- Generalized properly efficient solutions of vector optimization problems
- Steepest descent methods for multicriteria optimization.
- Descent algorithm for nonsmooth stochastic multiobjective optimization
- On min-norm and min-max methods of multi-objective optimization
- Average cost temporal-difference learning
- Finding intrinsic rewards by embodied evolution and constrained reinforcement learning
- Multi-objective Reinforcement Learning through Continuous Pareto Manifold Approximation
- A Survey of Multi-Objective Sequential Decision-Making
- Scenarios and Policy Aggregation in Optimization Under Uncertainty
- Sequential Approximate Multiobjective Optimization Using Computational Intelligence
- Multivariate stochastic approximation using a simultaneous perturbation gradient approximation
- Simulation-based optimization of Markov reward processes
- Interactive bundle-based method for nondifferentiable multiobjeective optimization: nimbus§
- A general reinforcement learning algorithm that masters chess, shogi, and Go through self-play
- Computing Convex Coverage Sets for Faster Multi-objective Coordination
