Efficient Multi-objective Reinforcement Learning via Multiple-gradient Descent with Iteratively Discovered Weight-Vector Sets.
Yongcan CaoHuixin ZhanPublished in: J. Artif. Intell. Res. (2021)
Keyphrases
- multi objective
- reinforcement learning
- weight vector
- objective function
- loss function
- stochastic gradient descent
- optimization algorithm
- multi objective optimization
- learning rate
- genetic algorithm
- feature space
- pairwise
- image retrieval
- evolutionary algorithm
- decision trees
- nearest neighbor
- linear combination
- image processing
- learning algorithm
- genetic algorithm ga
- information content
- machine learning