Zobrazeno 1 - 10
of 36
pro vyhledávání: '"Asadi, Kavosh"'
We focus on the task of learning the value function in the reinforcement learning (RL) setting. This task is often solved by updating a pair of online and target networks while ensuring that the parameters of these two networks are equivalent. We pro
Externí odkaz:
http://arxiv.org/abs/2406.01838
Autor:
Liu, Zuxin, Zhang, Jesse, Asadi, Kavosh, Liu, Yao, Zhao, Ding, Sabach, Shoham, Fakoor, Rasool
The full potential of large pretrained models remains largely untapped in control domains like robotics. This is mainly because of the scarcity of data and the computational challenges associated with training or fine-tuning these large models for su
Externí odkaz:
http://arxiv.org/abs/2310.05905
We focus on the task of approximating the optimal value function in deep reinforcement learning. This iterative process is comprised of solving a sequence of optimization problems where the loss function changes per iteration. The common approach to
Externí odkaz:
http://arxiv.org/abs/2306.17833
We study the convergence behavior of the celebrated temporal-difference (TD) learning algorithm. By looking at the algorithm through the lens of optimization, we first argue that TD can be viewed as an iterative optimization algorithm where the funct
Externí odkaz:
http://arxiv.org/abs/2306.17750
We study the action generalization ability of deep Q-learning in discrete action spaces. Generalization is crucial for efficient reinforcement learning (RL) because it allows agents to use knowledge learned from past experiences on new tasks. But whi
Externí odkaz:
http://arxiv.org/abs/2205.05588
Autor:
Asadi, Kavosh, Fakoor, Rasool, Gottesman, Omer, Kim, Taesup, Littman, Michael L., Smola, Alexander J.
Deep reinforcement learning algorithms often use two networks for value function optimization: an online network, and a target network that tracks the online network with some delay. Using two separate networks enables the agent to hedge against issu
Externí odkaz:
http://arxiv.org/abs/2112.05848
Autor:
Gottesman, Omer, Asadi, Kavosh, Allen, Cameron, Lobel, Sam, Konidaris, George, Littman, Michael
Principled decision-making in continuous state--action spaces is impossible without some assumptions. A common approach is to assume Lipschitz continuity of the Q-function. We show that, unfortunately, this property fails to hold in many typical doma
Externí odkaz:
http://arxiv.org/abs/2110.12276
Fluid human-agent communication is essential for the future of human-in-the-loop reinforcement learning. An agent must respond appropriately to feedback from its human trainer even before they have significant experience working together. Therefore,
Externí odkaz:
http://arxiv.org/abs/2109.07054
Reliant on too many experiments to learn good actions, current Reinforcement Learning (RL) algorithms have limited applicability in real-world settings, which can be too expensive to allow exploration. We propose an algorithm for batch RL, where effe
Externí odkaz:
http://arxiv.org/abs/2102.09225
Can simple algorithms with a good representation solve challenging reinforcement learning problems? In this work, we answer this question in the affirmative, where we take "simple learning algorithm" to be tabular Q-Learning, the "good representation
Externí odkaz:
http://arxiv.org/abs/2002.05518