Zobrazeno 1 - 10
of 465
pro vyhledávání: '"A Rakhsha"'
The Value Iteration (VI) algorithm is an iterative procedure to compute the value function of a Markov decision process, and is the basis of many reinforcement learning (RL) algorithms as well. As the error convergence rate of VI as a function of ite
Externí odkaz:
http://arxiv.org/abs/2407.10454
Long-horizon tasks, which have a large discount factor, pose a challenge for most conventional reinforcement learning (RL) algorithms. Algorithms such as Value Iteration and Temporal Difference (TD) learning have a slow convergence rate and become in
Externí odkaz:
http://arxiv.org/abs/2407.08803
Autor:
Salavatidezfouli, Sajad, Rakhsha, Saeid, Sheidani, Armin, Stabile, Giovanni, Rozza, Gianluigi
This paper aims to comprehensively investigate the efficacy of various Model Order Reduction (MOR) and deep learning techniques in predicting heat transfer in a pulsed jet impinging on a concave surface. Expanding on the previous experimental and num
Externí odkaz:
http://arxiv.org/abs/2402.10641
We introduce new planning and reinforcement learning algorithms for discounted MDPs that utilize an approximate model of the environment to accelerate the convergence of the value function. Inspired by the splitting approach in numerical linear algeb
Externí odkaz:
http://arxiv.org/abs/2211.13937
Autor:
Rakhsha, Amirhossein, Eslami, Reza, Yang, Xiaoxuan, Noor, Navid, Ismail, Fatma M., Abdellah, Ahmed M., Soleymani, Leyla, Higgins, Drew
Publikováno v:
In Nano Energy January 2025 133
Autor:
Navid Noor, Thomas Baker, Hyejin Lee, Elliot Evans, Shayan Angizi, Jeffrey Daniel Henderson, Amirhossein Rakhsha, Drew Higgins
Publikováno v:
ACS Omega, Vol 9, Iss 9, Pp 10080-10089 (2024)
Externí odkaz:
https://doaj.org/article/e319af2f7d3143958fd6560324f769fb
Autor:
Ahmed M. Abdellah, Fatma Ismail, Oliver W. Siig, Jie Yang, Carmen M. Andrei, Liza-Anastasia DiCecco, Amirhossein Rakhsha, Kholoud E. Salem, Kathryn Grandfield, Nabil Bassim, Robert Black, Georg Kastlunger, Leyla Soleymani, Drew Higgins
Publikováno v:
Nature Communications, Vol 15, Iss 1, Pp 1-15 (2024)
Abstract Electrochemical conversion of CO2 offers a sustainable route for producing fuels and chemicals. Pd-based catalysts are effective for converting CO2 into formate at low overpotentials and CO/H2 at high overpotentials, while undergoing poorly
Externí odkaz:
https://doaj.org/article/d1c7a154f14b4fbe8002d7f284550e96
Autor:
Keyvan Eghbal, Iman Ahrari, Fazlollah Kamrani, Somayeh Mohamamdi, Arash Saffarian, Mohamad Jamali, Abbas Rakhsha, Reza Taheri, Abdolkarim Rahmanian, Meisam Eqbal
Publikováno v:
Asian Journal of Surgery, Vol 46, Iss 9, Pp 3760-3765 (2023)
Purpose: Multilevel anterior cervical discectomy and fusion (ACDF), especially two-level ACDF, has been usually performed in the cervical degenerative disease, and the incidence rate of complications is controversial. This study aimed to compare the
Externí odkaz:
https://doaj.org/article/a690dca9d39b4ab18195888f80cb52d7
Reward Poisoning in Reinforcement Learning: Attacks Against Unknown Learners in Unknown Environments
We study black-box reward poisoning attacks against reinforcement learning (RL), in which an adversary aims to manipulate the rewards to mislead a sequence of RL agents with unknown algorithms to learn a nefarious policy in an environment unknown to
Externí odkaz:
http://arxiv.org/abs/2102.08492
Autor:
Faghihi Moghaddam, Fahimeh, Bakhshandeh, Mohsen, Mofid, Bahram, Sahinbas, Hüseyin, Faeghi, Fariborz, Mirzaei, Hamidreza, Rakhsha, Afshin, Yousefi Kashi, Amir Shahram, Sadeghi, Reza, Mahdavi, Arash
Publikováno v:
In Journal of Thermal Biology February 2024 120