Zobrazeno 1 - 10
of 80
pro vyhledávání: '"Machado, Marlos C."'
Deep neural networks can struggle to learn continually in the face of non-stationarity. This phenomenon is known as loss of plasticity. In this paper, we identify underlying principles that lead to plastic algorithms. In particular, we provide theore
Externí odkaz:
http://arxiv.org/abs/2410.20634
Publikováno v:
Reinforcement Learning Journal, vol. 1, no. 1, 2024
The recency heuristic in reinforcement learning is the assumption that stimuli that occurred closer in time to an acquired reward should be more heavily reinforced. The recency heuristic is one of the key assumptions made by TD($\lambda$), which rein
Externí odkaz:
http://arxiv.org/abs/2406.12284
Autor:
Lewandowski, Alex, Bortkiewicz, Michał, Kumar, Saurabh, György, András, Schuurmans, Dale, Ostaszewski, Mateusz, Machado, Marlos C.
Loss of plasticity is a phenomenon where neural networks can become more difficult to train over the course of learning. Continual learning algorithms seek to mitigate this effect by sustaining good performance while maintaining network trainability.
Externí odkaz:
http://arxiv.org/abs/2406.06811
Multistep returns, such as $n$-step returns and $\lambda$-returns, are commonly used to improve the sample efficiency of reinforcement learning (RL) methods. The variance of the multistep returns becomes the limiting factor in their length; looking t
Externí odkaz:
http://arxiv.org/abs/2402.03903
Autor:
Janjua, Muhammad Kamran, Shah, Haseeb, White, Martha, Miahi, Erfan, Machado, Marlos C., White, Adam
Publikováno v:
Machine Learning (2023): 1-31
In this paper we investigate the use of reinforcement-learning based prediction approaches for a real drinking-water treatment plant. Developing such a prediction system is a critical step on the path to optimizing and automating water treatment. Bef
Externí odkaz:
http://arxiv.org/abs/2312.01624
Reinforcement learning (RL) agents make decisions using nothing but observations from the environment, and consequently, heavily rely on the representations of those observations. Though some recent breakthroughs have used vector-based categorical re
Externí odkaz:
http://arxiv.org/abs/2312.01203
Loss of plasticity is a phenomenon in which neural networks lose their ability to learn from new experience. Despite being empirically observed in several problem settings, little is understood about the mechanisms that lead to loss of plasticity. In
Externí odkaz:
http://arxiv.org/abs/2312.00246
In this paper we investigate transformer architectures designed for partially observable online reinforcement learning. The self-attention mechanism in the transformer architecture is capable of capturing long-range dependencies and it is the main re
Externí odkaz:
http://arxiv.org/abs/2310.15719
The ability to learn good representations of states is essential for solving large reinforcement learning problems, where exploration, generalization, and transfer are particularly challenging. The Laplacian representation is a promising approach to
Externí odkaz:
http://arxiv.org/abs/2310.10833
The ability to learn continually is essential in a complex and changing world. In this paper, we characterize the behavior of canonical value-based deep reinforcement learning (RL) approaches under varying degrees of non-stationarity. In particular,
Externí odkaz:
http://arxiv.org/abs/2303.07507