Results 31 to 40 of about 6,292,535 (287)

Implementing action mask in proximal policy optimization (PPO) algorithm

ICT Express, 2020
The proximal policy optimization (PPO) algorithm is a promising algorithm in reinforcement learning. In this paper, we propose to add an action mask in the PPO algorithm. The mask indicates whether an action is valid or invalid for each state. Simulation
Cheng-Yen Tang, Chien-Hung Liu, Woei-Kae Chen, Shingchern D. You   +3 more
doaj   +1 more source

Neural Proximal/Trust Region Policy Optimization Attains Globally Optimal Policy [PDF]

, 2023
Proximal policy optimization and trust region policy optimization (PPO and TRPO) with actor and critic parametrized by neural networks achieve significant empirical success in deep reinforcement learning.
Liu, Boyi, Cai, Qi, Yang, Zhuoran, Wang, Zhaoran   +3 more
core   +1 more source

Joint action loss for proximal policy optimization

CoRR, 2023
PPO (Proximal Policy Optimization) is a state-of-the-art policy gradient algorithm that has been successfully applied to complex computer games such as Dota 2 and Honor of Kings. In these environments, an agent makes compound actions consisting of multiple sub-actions. PPO uses clipping to restrict policy updates.
Xiulei Song, Yizhao Jin, Greg G. Slabaugh, Simon Lucas 0001   +3 more
openaire   +2 more sources

Research on Reactive Power Margin Optimization of Doubly-Fed Induction Generator Wind Farms Based on Improved Proximal Policy Optimization Algorithm

发电技术
ObjectivesTo mitigate voltage fluctuations at the grid-connected points of doubly-fed induction generator wind farms while enhancing their reactive power margin, thereby strengthening their transient voltage support capability, a reactive power margin ...
YU Da, WU Zhitao, CHEN Ziwen, LIN Xu, WANG Weihua, LIU Chongru   +5 more
doaj   +1 more source

Minimax and entropic proximal policy optimization [PDF]

, 2023
First-order gradient descent is to date the most commonly used optimization method for training deep neural networks, especially for networks with shared parameters, or recurrent neural networks (RNNs).
Song, Yunlong
core   +1 more source

On Proximal Policy Optimization's Heavy-tailed Gradients

CoRR, 2021
Modern policy gradient algorithms such as Proximal Policy Optimization (PPO) rely on an arsenal of heuristics, including loss clipping and gradient clipping, to ensure successful learning. These heuristics are reminiscent of techniques from robust statistics, commonly used for estimation in outlier-rich (``heavy-tailed'') regimes.
Saurabh Garg, Joshua Zhanson, Emilio Parisotto, Adarsh Prasad, J. Zico Kolter, Sivaraman Balakrishnan, Zachary C. Lipton, Ruslan Salakhutdinov, Pradeep Ravikumar   +8 more
openaire   +4 more sources

Meta Proximal Policy Optimization for Cooperative Multi-Agent Continuous Control [PDF]

, 2022
In this thesis we propose Multi-Agent Proxy Proximal Policy Optimization (MA3PO), a novel multi-agent deep reinforcement learning algorithm that tackles the challenge of cooperative continuous multi-agent control.
Fang, Boli
core   +1 more source

A dual-arm cooperative control method based on improved proximal policy optimization

Journal of King Saud University: Computer and Information Sciences
To address the challenges of high-dimensional control in dual-arm collaborative tasks, the complexity of multi-stage decision-making, and the limitations of traditional Proximal Policy Optimization (PPO) algorithms due to their single constraint ...
Man Su, Qingni Yuan, Pengju Qu, Chao Wang, Yinjiang Zhou   +4 more
doaj   +1 more source

Early Clinical, Imaging, and Pathological Characteristics of SRPK3/TTN‐Digenic Myopathy

Annals of Clinical and Translational Neurology, EarlyView.
ABSTRACT Objective SRPK3/TTN‐digenic myopathy was recently established as a skeletal muscle myopathy caused by digenic inheritance. This study characterizes the early clinical presentation of SRPK3/TTN‐digenic myopathy in one previously reported and seven newly identified pediatric patients.
Rotem Orbach, Sandra Donkervoort, Carola Hedberg‐Oldfors, Giovanni Baranello, Dimah Saade, Precilla D'Souza, Ruchee Patel, Eva Michael, A. Reghan Foley, Diana Bharucha‐Goebel, S. Jin Haugland, Meghan McAnally, Omer Abdul Hamid, Katherine Chao, Ellen F. Macnamara, Alan H. Beggs, Anna Sarkozy, Juliane Mueller, Steven A. Moore, Richard S. Finkel, Cynthia J. Tifft, Francesco Muntoni, Anders Oldfors, Carsten G. Bönnemann   +23 more
wiley   +1 more source

Reparameterization Proximal Policy Optimization

CoRR
By leveraging differentiable dynamics, Reparameterization Policy Gradient (RPG) achieves high sample efficiency. However, current approaches are hindered by two critical limitations: the under-utilization of computationally expensive dynamics Jacobians and inherent training instability. While sample reuse offers a remedy for under-utilization, no prior
Hai Zhong, Xun Wang 0013, Zhuoran Li, Longbo Huang   +3 more
openaire   +2 more sources