Finite-Sample Analysis for Decentralized Batch Multiagent Reinforcement Learning With Networked Agents

Zhang, Kaiqing; Yang, Zhuoran; Liu, Han; Zhang, Tong; Başar, Tamer

doi:10.1109/tac.2021.3049345

Cited by 46 publications

(58 citation statements)

References 48 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is stated that immigrant students experience negativities such as loneliness or tendency to violence due to their inability to express themselves in social environments. Regarding school attendance problem (Zhang and Basar, 2018) stated that the lack of legislation for refugee students prevents the measures that can be taken against these students. It has been determined that the problem arising from the lack of legislation is the problem of absenteeism.…”

Section: Discussionmentioning

confidence: 99%

Duties of educators and administrators in adapting immigrant students to school

Kabataş¹

2021

Int. J. Educ. Admin. Pol. Stud.

View full text Add to dashboard Cite

In this study, the roles of teachers and school administrators in the adaptation of immigrant students to the school organization were investigated. For this purpose, the opinions of teachers and school administrators were taken for the adaptation of immigrant students to the school organization. The interview form used in the research was created by the researcher with the help of field experts. Participants consist of schools with a large number of immigrant students. The data obtained using face to face interview method was analyzed by content analysis technique. The research data were themed as studies on the adaptation of immigrant students to school, the adaptation and academic problems of immigrant students at school, and the solution of the problems encountered in the adaptation of immigrant students to school. Solution suggestions were tried to be found. As a result, there are important problems in the education of immigrant students. It may be suggested to plan serious activities in order to support teachers with in-service trainings on the education of immigrant children, to monitor students' attendance at school, and to eliminate communication problems between immigrant students and other students.

show abstract

Section: Discussionmentioning

confidence: 99%

Duties of educators and administrators in adapting immigrant students to school

Kabataş¹

2021

Int. J. Educ. Admin. Pol. Stud.

View full text Add to dashboard Cite

show abstract

“…of any non-empty participant is called a coalition. The Shapley value can be used to calculate the profit distributed by participant i, as shown in formulas: (27) and (28). !…”

Section: Cooperative Game Model 1) Profit Calculation Modelmentioning

confidence: 99%

“…In order to solve the optimal quotation problem of thermal power companies under the multi-agent incomplete information game, the Multi-Agent Deep Deterministic Policy Gradient (MADDPG) algorithm based on the multiagent reinforcement learning method was proposed [28][29][30][31] . The neural network parameters are updated to simulate the bounded rational process of the game to ensure that the game process is close to reality.…”

Section: Introductionmentioning

confidence: 99%

Research on Bidding Strategy of Thermal Power Companies in Electricity Market Based on Multi-Agent Deep Deterministic Policy Gradient

et al. 2021

View full text Add to dashboard Cite

With the continuous improvement of new energy penetration in the power system, the price of the spot market of power frequently fluctuates greatly, which damages the income of a large number of thermal power enterprises. In order to lock in the profit, thermal power enterprises should turn the main target of profit to the medium and long-term power market. With the continuous advancement of the reform in China's power system, major changes have taken place in the medium and long-term power transactions, including the transaction target, organization method, clearing method and so on, so it is urgent to explore the quotation strategy of thermal power enterprises under the medium and long term market changes. Based on the theory of game equilibrium, this paper establishes non-cooperative game and cooperative game models between thermal power companies. Considering that the traditional reinforcement learning method is difficult to solve the multi-agent incomplete information game model, this paper uses the Multi-Agent Deep Deterministic Policy Gradient(MADDPG) algorithm to solve the above model. Finally, the validity of the proposed model is proved by a numerical example. The results show that, compared with other reinforcement learning algorithms, when solving the multi-agent incomplete information game model, the quotation obtained by MADDPG is more accurate, the revenue is increased by 5.2%, and the convergence time is reduced by 50%.In addition, this paper finds that in the medium and long-term power market, thermal power companies are more inclined to use physical retention methods to make profits. The greater the market power of thermal power companies, the greater the probability of physical retention. When low-cost thermal power companies retain more power, they will increase market clearing electricity prices and harm market efficiency. Regulators should focus on the market behavior of such thermal power companies.

show abstract

“…COMA (Foerster et al 2018) constructs a centralized critic and computes an agent-specific advantage function to derive a decentralized actor. FDMARL (Zhang et al 2018) has proposed a distributed learning approach for each agent to learn a global critic using its local reward and the transferred critic parameters from the networked neighboring agents. Because these models directly use the state or observation in constructing critic or actor networks, it is difficult to apply such models to a large-scale environment or transfer them to new environments.…”

Section: Learning-for-consensusmentioning

confidence: 99%

Multi-Agent Actor-Critic with Hierarchical Graph Attention Network

Ryu

Shin

Park

2020

AAAI

View full text Add to dashboard Cite

Most previous studies on multi-agent reinforcement learning focus on deriving decentralized and cooperative policies to maximize a common reward and rarely consider the transferability of trained policies to new tasks. This prevents such policies from being applied to more complex multi-agent tasks. To resolve these limitations, we propose a model that conducts both representation learning for multiple agents using hierarchical graph attention network and policy learning using multi-agent actor-critic. The hierarchical graph attention network is specially designed to model the hierarchical relationships among multiple agents that either cooperate or compete with each other to derive more advanced strategic policies. Two attention networks, the inter-agent and inter-group attention layers, are used to effectively model individual and group level interactions, respectively. The two attention networks have been proven to facilitate the transfer of learned policies to new tasks with different agent compositions and allow one to interpret the learned strategies. Empirically, we demonstrate that the proposed model outperforms existing methods in several mixed cooperative and competitive tasks.

show abstract

Finite-Sample Analysis for Decentralized Batch Multiagent Reinforcement Learning With Networked Agents

Cited by 46 publications

References 48 publications

Duties of educators and administrators in adapting immigrant students to school

Duties of educators and administrators in adapting immigrant students to school

Research on Bidding Strategy of Thermal Power Companies in Electricity Market Based on Multi-Agent Deep Deterministic Policy Gradient

Multi-Agent Actor-Critic with Hierarchical Graph Attention Network

Contact Info

Product

Resources

About