Pub Date : 2004-10-18DOI: 10.1093/ietisy/e88-d.5.1004
Hiroshi Osada, S. Fujita
We propose a reinforcement learning scheme called CHQ that could efficiently acquire appropriate policies under partially observable Markov decision processes (POMDP) involving probabilistic state transitions, that frequently occurs in multiagent systems in which each agent independently takes a probabilistic action based on a partial observation of the underlying environment. A key idea of CHQ is to extend the HQ-learning proposed by Wiering et al. in such a way that it could learn the activation order of the MDP subtasks as well as an appropriate policy under each MDP subtask. The quality of the proposed scheme is experimentally evaluated. The result of experiments implies that it can acquire a deterministic policy with sufficiently high success rate, even if the given task is POMDP with probabilistic state transitions.
{"title":"CHQ: a multi-agent reinforcement learning scheme for partially observable Markov decision processes","authors":"Hiroshi Osada, S. Fujita","doi":"10.1093/ietisy/e88-d.5.1004","DOIUrl":"https://doi.org/10.1093/ietisy/e88-d.5.1004","url":null,"abstract":"We propose a reinforcement learning scheme called CHQ that could efficiently acquire appropriate policies under partially observable Markov decision processes (POMDP) involving probabilistic state transitions, that frequently occurs in multiagent systems in which each agent independently takes a probabilistic action based on a partial observation of the underlying environment. A key idea of CHQ is to extend the HQ-learning proposed by Wiering et al. in such a way that it could learn the activation order of the MDP subtasks as well as an appropriate policy under each MDP subtask. The quality of the proposed scheme is experimentally evaluated. The result of experiments implies that it can acquire a deterministic policy with sufficiently high success rate, even if the given task is POMDP with probabilistic state transitions.","PeriodicalId":281008,"journal":{"name":"Proceedings. IEEE/WIC/ACM International Conference on Intelligent Agent Technology, 2004. (IAT 2004).","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130431477","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2004-09-20DOI: 10.1109/IAT.2004.1343015
S. Rubin, W. Dai
Interface agents have specialized roles interacting with human users and provide communication channels between external and internal worlds of underlying information systems or within a multiagent team. We discuss the roles and functions of a natural language enabled interface agent (NLEIA). We are particularly interested in integrating NLEIA with other agents within a multiagent project team. The impact of having NLEIA on the overall performance and the usability of a complex software system in general and multiagent team in particular are also discussed. The paper starts with the background technologies for natural language processing (NLP), which provide the implementation basis for NLEIA, followed with the description of roles and activities of NLEIA. Finally, several application examples demonstrating the feasibility of the proposed approach are presented.
{"title":"Natural language enabled interface agent","authors":"S. Rubin, W. Dai","doi":"10.1109/IAT.2004.1343015","DOIUrl":"https://doi.org/10.1109/IAT.2004.1343015","url":null,"abstract":"Interface agents have specialized roles interacting with human users and provide communication channels between external and internal worlds of underlying information systems or within a multiagent team. We discuss the roles and functions of a natural language enabled interface agent (NLEIA). We are particularly interested in integrating NLEIA with other agents within a multiagent project team. The impact of having NLEIA on the overall performance and the usability of a complex software system in general and multiagent team in particular are also discussed. The paper starts with the background technologies for natural language processing (NLP), which provide the implementation basis for NLEIA, followed with the description of roles and activities of NLEIA. Finally, several application examples demonstrating the feasibility of the proposed approach are presented.","PeriodicalId":281008,"journal":{"name":"Proceedings. IEEE/WIC/ACM International Conference on Intelligent Agent Technology, 2004. (IAT 2004).","volume":"141 ","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-09-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120868488","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2004-09-20DOI: 10.1109/IAT.2004.1342969
Liming Wang, Houkuan Huang, Yu-mei Chai
Speculative computation is based on abduction, and it can give result of decision in advance based on defaults under incomplete resource information. To achieve goals and reduce risks of decision, master agent in speculative computation makes efforts to obtain more information via negotiation. However, the speculative computation and negotiation have usually deadlines. In This work, framework extended of speculative computation with deadline is presented, and the further negotiation framework and negotiation algorithm under time constraints are presented, the algorithm is embedded in speculative computation. The experiments have proved the algorithm can improve the accuracy of result of speculative computation, reduce the risk of the result.
{"title":"Speculative computation with deadline and its resource negotiation under time constraints","authors":"Liming Wang, Houkuan Huang, Yu-mei Chai","doi":"10.1109/IAT.2004.1342969","DOIUrl":"https://doi.org/10.1109/IAT.2004.1342969","url":null,"abstract":"Speculative computation is based on abduction, and it can give result of decision in advance based on defaults under incomplete resource information. To achieve goals and reduce risks of decision, master agent in speculative computation makes efforts to obtain more information via negotiation. However, the speculative computation and negotiation have usually deadlines. In This work, framework extended of speculative computation with deadline is presented, and the further negotiation framework and negotiation algorithm under time constraints are presented, the algorithm is embedded in speculative computation. The experiments have proved the algorithm can improve the accuracy of result of speculative computation, reduce the risk of the result.","PeriodicalId":281008,"journal":{"name":"Proceedings. IEEE/WIC/ACM International Conference on Intelligent Agent Technology, 2004. (IAT 2004).","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-09-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125163205","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2004-09-20DOI: 10.1109/IAT.2004.1342985
Xiang Feng, D. Shuai
This work presents a new spring net approach for distributed problem solving in MAS, which is entirely different from the EN for TSP and can describe a variety of complicated social interactive behavior and autonomy of agents. The simulations of task allocation and resource assignment have shown the advantages of the proposed spring net approach for distributed problem solving in MAS.
{"title":"A new spring net approach to distributed problem solving in multi-agent systems","authors":"Xiang Feng, D. Shuai","doi":"10.1109/IAT.2004.1342985","DOIUrl":"https://doi.org/10.1109/IAT.2004.1342985","url":null,"abstract":"This work presents a new spring net approach for distributed problem solving in MAS, which is entirely different from the EN for TSP and can describe a variety of complicated social interactive behavior and autonomy of agents. The simulations of task allocation and resource assignment have shown the advantages of the proposed spring net approach for distributed problem solving in MAS.","PeriodicalId":281008,"journal":{"name":"Proceedings. IEEE/WIC/ACM International Conference on Intelligent Agent Technology, 2004. (IAT 2004).","volume":"197 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-09-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126026340","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2004-09-20DOI: 10.1109/IAT.2004.1342967
Longbing Cao, Jiaqi Wang, Li Lin, Chengqi Zhang
Traders and researchers in stock marketing often hold some private trading strategies. Evaluation and optimization of their strategies is a great benefit to them before they take any risk in realistic trading. We build an agent services-driven infrastructure: F-TRADE. It supports online plug in, iterative back-test, and recommendation of trading strategies. We propose agent services-driven approach for building the above automated enterprise infrastructure. Description, directory and mediation of agent services are discussed. System structure of the agent services-based F-TRADE is also discussed. F-TRADE has been an online test platform for research and application of multi-agent technology, and data mining in stock markets.
{"title":"Agent services-based infrastructure for online assessment of trading strategies","authors":"Longbing Cao, Jiaqi Wang, Li Lin, Chengqi Zhang","doi":"10.1109/IAT.2004.1342967","DOIUrl":"https://doi.org/10.1109/IAT.2004.1342967","url":null,"abstract":"Traders and researchers in stock marketing often hold some private trading strategies. Evaluation and optimization of their strategies is a great benefit to them before they take any risk in realistic trading. We build an agent services-driven infrastructure: F-TRADE. It supports online plug in, iterative back-test, and recommendation of trading strategies. We propose agent services-driven approach for building the above automated enterprise infrastructure. Description, directory and mediation of agent services are discussed. System structure of the agent services-based F-TRADE is also discussed. F-TRADE has been an online test platform for research and application of multi-agent technology, and data mining in stock markets.","PeriodicalId":281008,"journal":{"name":"Proceedings. IEEE/WIC/ACM International Conference on Intelligent Agent Technology, 2004. (IAT 2004).","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-09-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115121927","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2004-09-20DOI: 10.1109/IAT.2004.1343008
R. Howard, L. Kerschberg
The concept of automating Web services, specifically the brokering activities, is an active research topic. We need a comprehensive and overarching framework that seamlessly incorporates intelligent middleware agents within the context of workflow management, and addresses the issues related to virtual organizations. The goal is to add semantics to Web services to endow them with capabilities currently lacking in the literature, but necessary for their successful deployment in future systems. This paper discusses how the knowledge-based dynamic semantic Web services (KDSWS) framework, interoperating with the knowledge sifter architecture, can be used to dynamically broker semantic Web services. The functional agent services architecture of the KDSWS Framework contains intelligent middleware agents to enable these dynamic operations. In particular, the specification of the roles of the intelligent middleware agents in the brokering is presented.
{"title":"Brokering semantic Web services via intelligent middleware agents within a knowledge-based framework","authors":"R. Howard, L. Kerschberg","doi":"10.1109/IAT.2004.1343008","DOIUrl":"https://doi.org/10.1109/IAT.2004.1343008","url":null,"abstract":"The concept of automating Web services, specifically the brokering activities, is an active research topic. We need a comprehensive and overarching framework that seamlessly incorporates intelligent middleware agents within the context of workflow management, and addresses the issues related to virtual organizations. The goal is to add semantics to Web services to endow them with capabilities currently lacking in the literature, but necessary for their successful deployment in future systems. This paper discusses how the knowledge-based dynamic semantic Web services (KDSWS) framework, interoperating with the knowledge sifter architecture, can be used to dynamically broker semantic Web services. The functional agent services architecture of the KDSWS Framework contains intelligent middleware agents to enable these dynamic operations. In particular, the specification of the roles of the intelligent middleware agents in the brokering is presented.","PeriodicalId":281008,"journal":{"name":"Proceedings. IEEE/WIC/ACM International Conference on Intelligent Agent Technology, 2004. (IAT 2004).","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-09-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128963202","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This work investigates the effectiveness of an agent based trading system. The system developed employs a simple genetic algorithm to evolve an optimized trading approach for every agent, with their trading decisions based on a range of technical indicators generating trading signals. Their trading pattern follows a simple fitness function of maximizing net assets for every evolutionary cycle. Their performance is analyzed compared to market movements as represented by its index, as well as investment funds run by human professionals to establish a relative measure of success. The results show that the developed system performs comparably to its human counterparts across different market environments, despite these agents being rather primitive in nature. Future forthcoming work refines and explores the potential of this approach further.
{"title":"Agent based genetic algorithm employing financial technical analysis for making trading decisions using historical equity market data","authors":"C. Schoreels, B. Logan, J. Garibaldi","doi":"10.1109/IAT.2004.23","DOIUrl":"https://doi.org/10.1109/IAT.2004.23","url":null,"abstract":"This work investigates the effectiveness of an agent based trading system. The system developed employs a simple genetic algorithm to evolve an optimized trading approach for every agent, with their trading decisions based on a range of technical indicators generating trading signals. Their trading pattern follows a simple fitness function of maximizing net assets for every evolutionary cycle. Their performance is analyzed compared to market movements as represented by its index, as well as investment funds run by human professionals to establish a relative measure of success. The results show that the developed system performs comparably to its human counterparts across different market environments, despite these agents being rather primitive in nature. Future forthcoming work refines and explores the potential of this approach further.","PeriodicalId":281008,"journal":{"name":"Proceedings. IEEE/WIC/ACM International Conference on Intelligent Agent Technology, 2004. (IAT 2004).","volume":"57 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-09-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116471506","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2004-09-20DOI: 10.1109/IAT.2004.1343011
J. Just, M. Cornwell, M. Huhns
Ad hoc cross-agency teams are often needed to deal with actual, imminent, or potential crises involving multiple geographic/political jurisdictions or requiring coordinated expertise from organizations with different responsibilities. Our initial ontology-based, policy-driven version of personal agents and agent-based work flows and Web services facilitates the establishment of such teams. The agents and services use and augment the existing email and office automation applications of an organization. Preliminary results indicate a potential reduction of two orders of magnitude in the time needed to form a team.
{"title":"Agents for establishing ad hoc cross-organizational teams","authors":"J. Just, M. Cornwell, M. Huhns","doi":"10.1109/IAT.2004.1343011","DOIUrl":"https://doi.org/10.1109/IAT.2004.1343011","url":null,"abstract":"Ad hoc cross-agency teams are often needed to deal with actual, imminent, or potential crises involving multiple geographic/political jurisdictions or requiring coordinated expertise from organizations with different responsibilities. Our initial ontology-based, policy-driven version of personal agents and agent-based work flows and Web services facilitates the establishment of such teams. The agents and services use and augment the existing email and office automation applications of an organization. Preliminary results indicate a potential reduction of two orders of magnitude in the time needed to form a team.","PeriodicalId":281008,"journal":{"name":"Proceedings. IEEE/WIC/ACM International Conference on Intelligent Agent Technology, 2004. (IAT 2004).","volume":"13 50 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-09-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114731277","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2004-09-20DOI: 10.1109/IAT.2004.1342939
Sherief Abdallah, V. Lesser
The coalition formation problem has received a considerable amount of attention in recent years. In this work we present a novel distributed algorithm that returns a solution in polynomial time and the quality of the returned solution increases as agents gain more experience. Our solution utilizes an underlying organization to guide the coalition formation process. We use reinforcement learning techniques to optimize decisions made locally by agents in the organization. Experimental results are presented, showing the potential of our approach.
{"title":"Organization-based cooperative coalition formation","authors":"Sherief Abdallah, V. Lesser","doi":"10.1109/IAT.2004.1342939","DOIUrl":"https://doi.org/10.1109/IAT.2004.1342939","url":null,"abstract":"The coalition formation problem has received a considerable amount of attention in recent years. In this work we present a novel distributed algorithm that returns a solution in polynomial time and the quality of the returned solution increases as agents gain more experience. Our solution utilizes an underlying organization to guide the coalition formation process. We use reinforcement learning techniques to optimize decisions made locally by agents in the organization. Experimental results are presented, showing the potential of our approach.","PeriodicalId":281008,"journal":{"name":"Proceedings. IEEE/WIC/ACM International Conference on Intelligent Agent Technology, 2004. (IAT 2004).","volume":"62 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-09-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114743392","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2004-09-20DOI: 10.1109/IAT.2004.1343010
Ben Choi, R. Dhawan
The future of computing is moving from individual processing units to communities of self organizing agents. In This work we propose an agent and network based architecture for parallel and distributed computing called agent space architecture. Our architecture builds upon the notions of agent and object space and utilizes multicast networks. The building blocks for our proposed architecture consist of an active processing unit called agent, a shared place for communication called space, and a communication medium called multicast network. One unique feature of our architecture is that we extend the concept of object space to become an active space. Our active space functions as a rendezvous, a repository, a cache, a responder, a notifier, and a manager of its own resources. The organization of our architecture is as general as network topology. Any number of agents, spaces, or networks can be added to achieve high performance. It is as scalable as Ethernet and adding agents or spaces is as easy as plug and play. High availability and fault tolerance is achieved through multiple agents, spaces, and networks. All these features are particularly beneficial for challenging applications such as search engine, which is used as a test case to implement and to test our proposed architecture.
{"title":"Agent space architecture for search engines","authors":"Ben Choi, R. Dhawan","doi":"10.1109/IAT.2004.1343010","DOIUrl":"https://doi.org/10.1109/IAT.2004.1343010","url":null,"abstract":"The future of computing is moving from individual processing units to communities of self organizing agents. In This work we propose an agent and network based architecture for parallel and distributed computing called agent space architecture. Our architecture builds upon the notions of agent and object space and utilizes multicast networks. The building blocks for our proposed architecture consist of an active processing unit called agent, a shared place for communication called space, and a communication medium called multicast network. One unique feature of our architecture is that we extend the concept of object space to become an active space. Our active space functions as a rendezvous, a repository, a cache, a responder, a notifier, and a manager of its own resources. The organization of our architecture is as general as network topology. Any number of agents, spaces, or networks can be added to achieve high performance. It is as scalable as Ethernet and adding agents or spaces is as easy as plug and play. High availability and fault tolerance is achieved through multiple agents, spaces, and networks. All these features are particularly beneficial for challenging applications such as search engine, which is used as a test case to implement and to test our proposed architecture.","PeriodicalId":281008,"journal":{"name":"Proceedings. IEEE/WIC/ACM International Conference on Intelligent Agent Technology, 2004. (IAT 2004).","volume":"2019 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-09-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121455341","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：