Pub Date : 2009-01-01DOI: 10.2478/v10143-010-0004-z
Galina Merkuryeva, O. Vecherinska, J. Hatem
Statistical Input Data Analysis for Supply Chain Simulation Stochastic simulation models utilize probability distributions to represent a multitude of randomly occurring events. Theoretical distributions are used to represent empirical data because they help smooth data irregularities that may exist due to values missed during the data collection period. The incompatibility between specific characteristics of the theoretical distribution and assumptions of simulation and mathematical calculus present an actual problem in supply chains. The paper is based on the analysis of mentioned contradictions. Different approaches to deal with theoretical probability distributions in supply chains are described in the paper.
{"title":"Statistical Input Data Analysis for Supply Chain Simulation","authors":"Galina Merkuryeva, O. Vecherinska, J. Hatem","doi":"10.2478/v10143-010-0004-z","DOIUrl":"https://doi.org/10.2478/v10143-010-0004-z","url":null,"abstract":"Statistical Input Data Analysis for Supply Chain Simulation Stochastic simulation models utilize probability distributions to represent a multitude of randomly occurring events. Theoretical distributions are used to represent empirical data because they help smooth data irregularities that may exist due to values missed during the data collection period. The incompatibility between specific characteristics of the theoretical distribution and assumptions of simulation and mathematical calculus present an actual problem in supply chains. The paper is based on the analysis of mentioned contradictions. Different approaches to deal with theoretical probability distributions in supply chains are described in the paper.","PeriodicalId":211660,"journal":{"name":"Sci. J. Riga Tech. Univ. Ser. Comput. Sci.","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128107630","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 : 2009-01-01DOI: 10.2478/v10143-010-0001-2
L. Napalkova
Hybridisation of Evolutionary Algorithms for Solving Multi-Objective Simulation Optimisation Problems The paper presents a taxonomic analysis of existing hybrid multi-objective evolutionary algorithms aimed at solving multi-objective simulation optimisation problems. For that, the properties of evolutionary algorithms and the requirements made to solving the problem considered are determined. Finally, a combination of the properties, which allows one to increase the approximation accuracy of the Pareto-optimal front at relatively low computational costs, is revealed.
{"title":"Hybridisation of Evolutionary Algorithms for Solving Multi-Objective Simulation Optimisation Problems","authors":"L. Napalkova","doi":"10.2478/v10143-010-0001-2","DOIUrl":"https://doi.org/10.2478/v10143-010-0001-2","url":null,"abstract":"Hybridisation of Evolutionary Algorithms for Solving Multi-Objective Simulation Optimisation Problems The paper presents a taxonomic analysis of existing hybrid multi-objective evolutionary algorithms aimed at solving multi-objective simulation optimisation problems. For that, the properties of evolutionary algorithms and the requirements made to solving the problem considered are determined. Finally, a combination of the properties, which allows one to increase the approximation accuracy of the Pareto-optimal front at relatively low computational costs, is revealed.","PeriodicalId":211660,"journal":{"name":"Sci. J. Riga Tech. Univ. Ser. Comput. Sci.","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116975095","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 : 2009-01-01DOI: 10.2478/v10143-010-0003-0
Vladimirs Lajevskis, Pjotrs Dorogovs, A. Romānovs
It Security System Development for State Institution Nowadays the work efficiency and proficiency of public and private sector substantially depend on information technologies. In connection with the large volume of information stored in automated systems throughout the public sector, special attention should be paid to information security and security of actives. The main goals and tasks of information security, frameworks of IT security system are defined and analyzed in the paper; the standard of IT security system development is investigated, as well as international standards significance and guidance application within the development of IT security system. On the bases of research, the features of the State institutions are defined, which are necessary to take into consideration during the development of IT security system. The primary tasks of IT security system development are emphasized and the technique of development of IT security systems for State institutions is worked out.
{"title":"It Security System Development for State Institution","authors":"Vladimirs Lajevskis, Pjotrs Dorogovs, A. Romānovs","doi":"10.2478/v10143-010-0003-0","DOIUrl":"https://doi.org/10.2478/v10143-010-0003-0","url":null,"abstract":"It Security System Development for State Institution Nowadays the work efficiency and proficiency of public and private sector substantially depend on information technologies. In connection with the large volume of information stored in automated systems throughout the public sector, special attention should be paid to information security and security of actives. The main goals and tasks of information security, frameworks of IT security system are defined and analyzed in the paper; the standard of IT security system development is investigated, as well as international standards significance and guidance application within the development of IT security system. On the bases of research, the features of the State institutions are defined, which are necessary to take into consideration during the development of IT security system. The primary tasks of IT security system development are emphasized and the technique of development of IT security systems for State institutions is worked out.","PeriodicalId":211660,"journal":{"name":"Sci. J. Riga Tech. Univ. Ser. Comput. Sci.","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129986217","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 : 2009-01-01DOI: 10.2478/v10143-010-0002-1
A. Tambovcevs, Y. Merkuryev
Analysis of ERP Systems Implementation in the Construction Enterprises Effective work requires integrated systems that can share access to a common data set. Integrated workplace organizations in the construction company are requiring enterprise resource planning (ERP)-type systems that, in turn, tie together all relevant logistic, facilities, human resource, financial, and project data into a single, shared database. Unfortunately, many ERP systems make things more complex for construction business. The reality is that few ERP systems are vertically integrated for use in construction, and most of them are very complicate and ultimately difficult to implement. In addition, suppliers of ERP systems, for example iScala, SAP, Oracle and etc., must work with consultants and integrators to provide normal systems job after implementation. The purpose of the research is to identify, investigate, analyze and systematize the factors that can influence creation and functioning of the ERP system in a company and create metodology of ERP system implementation. ERP system modules for construction enterprise were evaluated, benefits and risks of ERP system were summarized.
{"title":"Analysis of ERP Systems Implementation in the Construction Enterprises","authors":"A. Tambovcevs, Y. Merkuryev","doi":"10.2478/v10143-010-0002-1","DOIUrl":"https://doi.org/10.2478/v10143-010-0002-1","url":null,"abstract":"Analysis of ERP Systems Implementation in the Construction Enterprises Effective work requires integrated systems that can share access to a common data set. Integrated workplace organizations in the construction company are requiring enterprise resource planning (ERP)-type systems that, in turn, tie together all relevant logistic, facilities, human resource, financial, and project data into a single, shared database. Unfortunately, many ERP systems make things more complex for construction business. The reality is that few ERP systems are vertically integrated for use in construction, and most of them are very complicate and ultimately difficult to implement. In addition, suppliers of ERP systems, for example iScala, SAP, Oracle and etc., must work with consultants and integrators to provide normal systems job after implementation. The purpose of the research is to identify, investigate, analyze and systematize the factors that can influence creation and functioning of the ERP system in a company and create metodology of ERP system implementation. ERP system modules for construction enterprise were evaluated, benefits and risks of ERP system were summarized.","PeriodicalId":211660,"journal":{"name":"Sci. J. Riga Tech. Univ. Ser. Comput. Sci.","volume":"106 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134639944","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 : 2009-01-01DOI: 10.2478/v10143-010-0008-8
S. Berzisa
The Baseline Configuration of Project Management Information System The purpose of project management information systems is to facilitate development and management of projects as well as to provide operational project information. Project management information systems need to be configured according to project management requirements during their deployment. In order to reduce complexity of this task, a baseline configuration can be used as a starting point of the configuration process. The objective of this paper is to establish the baseline configuration of the project management information system. The baseline configuration is developed according a set of typical project management requirements These requirements are determined by summarizing several project management methodologies and investigations on evaluation of project management information systems. They are divided in 15 groups. A high level description of the baseline configuration of project management information system is defined on the basis of these requirements. It consists of the list of typical project management scenarios and workflows as well as the list of relevant concepts, their default values and business rules. The baseline configuration is compared with configurations used in selected project management information systems. The obtained results further will be used in the configuration process of project management information systems.
{"title":"The Baseline Configuration of Project Management Information System","authors":"S. Berzisa","doi":"10.2478/v10143-010-0008-8","DOIUrl":"https://doi.org/10.2478/v10143-010-0008-8","url":null,"abstract":"The Baseline Configuration of Project Management Information System The purpose of project management information systems is to facilitate development and management of projects as well as to provide operational project information. Project management information systems need to be configured according to project management requirements during their deployment. In order to reduce complexity of this task, a baseline configuration can be used as a starting point of the configuration process. The objective of this paper is to establish the baseline configuration of the project management information system. The baseline configuration is developed according a set of typical project management requirements These requirements are determined by summarizing several project management methodologies and investigations on evaluation of project management information systems. They are divided in 15 groups. A high level description of the baseline configuration of project management information system is defined on the basis of these requirements. It consists of the list of typical project management scenarios and workflows as well as the list of relevant concepts, their default values and business rules. The baseline configuration is compared with configurations used in selected project management information systems. The obtained results further will be used in the configuration process of project management information systems.","PeriodicalId":211660,"journal":{"name":"Sci. J. Riga Tech. Univ. Ser. Comput. Sci.","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127735716","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 : 2009-01-01DOI: 10.2478/v10143-010-0005-y
Galina Merkuryeva, V. Bolshakovs
Simulation-Based Analysis of Fitness Landscape in Optimisation Simulation-based analysis of fitness landscape with application to optimisation problems is discussed in the paper. Methods of analysis of fitness landscapes and measures known in literature are reviewed. Procedure for simulation-based analysis of fitness landscape is introduced. Software prototype to perform this analysis is described. Case study for a vehicle scheduling problem with the time window constraints is given and demonstrates the main steps of fitness landscape analysis applied to simulation optimisation problem.
{"title":"Simulation-Based Analysis of Fitness Landscape in Optimisation","authors":"Galina Merkuryeva, V. Bolshakovs","doi":"10.2478/v10143-010-0005-y","DOIUrl":"https://doi.org/10.2478/v10143-010-0005-y","url":null,"abstract":"Simulation-Based Analysis of Fitness Landscape in Optimisation Simulation-based analysis of fitness landscape with application to optimisation problems is discussed in the paper. Methods of analysis of fitness landscapes and measures known in literature are reviewed. Procedure for simulation-based analysis of fitness landscape is introduced. Software prototype to perform this analysis is described. Case study for a vehicle scheduling problem with the time window constraints is given and demonstrates the main steps of fitness landscape analysis applied to simulation optimisation problem.","PeriodicalId":211660,"journal":{"name":"Sci. J. Riga Tech. Univ. Ser. Comput. Sci.","volume":"212 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122879988","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 : 2009-01-01DOI: 10.2478/v10143-010-0007-9
Arnis Lektauers
Multi-Agent Geosimulation of Urban Dynamics within the V-Devs Framework In recent years, a number of research efforts have focused on presenting solutions for modelling and simulation of urban systems using automata-based models allowing a transition from large scale, aggregate spatial representation in a static equilibrium to much finer scale disaggregate forms where dynamic processes are the primary focus of the simulation. In this paper a conceptual framework for integrating of agent-based models with cellular automata, geographic automata systems and geographic information systems in the context of modelling and simulation of urban systems is proposed. The proposed framework uses Visual Discrete Event System Specification (V-DEVS) that provides integrated modelling and simulation in a virtual interactive simulation environment, in a such way solving the existing problem that multi-agent systems and geographic information systems separately do not provide truly integrative modelling and simulation capabilities.
近年来，许多研究工作都集中在使用基于自动机的模型为城市系统的建模和模拟提供解决方案，允许从静态平衡中的大规模聚合空间表示过渡到更精细的尺度分解形式，其中动态过程是模拟的主要焦点。本文提出了在城市系统建模与仿真的背景下，将基于agent的模型与元胞自动机、地理自动机系统和地理信息系统相结合的概念框架。提出的框架采用可视化离散事件系统规范(Visual Discrete Event System Specification, V-DEVS)，在虚拟交互仿真环境中提供集成建模和仿真，从而解决了多智能体系统和地理信息系统各自无法提供真正集成建模和仿真能力的问题。
{"title":"Multi-Agent Geosimulation of Urban Dynamics within the V-Devs Framework","authors":"Arnis Lektauers","doi":"10.2478/v10143-010-0007-9","DOIUrl":"https://doi.org/10.2478/v10143-010-0007-9","url":null,"abstract":"Multi-Agent Geosimulation of Urban Dynamics within the V-Devs Framework In recent years, a number of research efforts have focused on presenting solutions for modelling and simulation of urban systems using automata-based models allowing a transition from large scale, aggregate spatial representation in a static equilibrium to much finer scale disaggregate forms where dynamic processes are the primary focus of the simulation. In this paper a conceptual framework for integrating of agent-based models with cellular automata, geographic automata systems and geographic information systems in the context of modelling and simulation of urban systems is proposed. The proposed framework uses Visual Discrete Event System Specification (V-DEVS) that provides integrated modelling and simulation in a virtual interactive simulation environment, in a such way solving the existing problem that multi-agent systems and geographic information systems separately do not provide truly integrative modelling and simulation capabilities.","PeriodicalId":211660,"journal":{"name":"Sci. J. Riga Tech. Univ. Ser. Comput. Sci.","volume":"13 4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126836901","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 : 2009-01-01DOI: 10.2478/v10143-010-0006-x
O. Soshko
Inventory Management in Multi Echelon Supply Chain using Sample Average Approximation An optimization model of multiechelon supply chain is presented in this paper. The decisions to be made are the amount of beer to be ordered in every echelon of supply chain in each echelon over the time horizon of one year. Since demand of the end customer is stochastic and presented by means of scenarios, the problem is solved by using sample average approximation method. This method uses only a subset of the scenarios, randomly sampled according to the distribution over scenarios, to represent the full scenario space. An important theoretical justification for this method is that as the number of scenarios sampled increases, the solution to the approximate problem converges to an optimal solution in the expected sense. The computational results are presented for two cases. First target level is chosen as a decision variable and then order size is chosen as a decision variable of the problem. The target level strategy is based on making inventory for each echelon; in its turn order strategy is based on determination of optimal order quantity, which is independent from scenarios. However target level strategy provides high service at low cost, but it offers less reality under uncertain demand than order strategy. Practical experiments on finding the optimal SAA parameters are presented in the paper and as well as the analysis of their impact on solution quality.
{"title":"Inventory Management in Multi Echelon Supply Chain using Sample Average Approximation","authors":"O. Soshko","doi":"10.2478/v10143-010-0006-x","DOIUrl":"https://doi.org/10.2478/v10143-010-0006-x","url":null,"abstract":"Inventory Management in Multi Echelon Supply Chain using Sample Average Approximation An optimization model of multiechelon supply chain is presented in this paper. The decisions to be made are the amount of beer to be ordered in every echelon of supply chain in each echelon over the time horizon of one year. Since demand of the end customer is stochastic and presented by means of scenarios, the problem is solved by using sample average approximation method. This method uses only a subset of the scenarios, randomly sampled according to the distribution over scenarios, to represent the full scenario space. An important theoretical justification for this method is that as the number of scenarios sampled increases, the solution to the approximate problem converges to an optimal solution in the expected sense. The computational results are presented for two cases. First target level is chosen as a decision variable and then order size is chosen as a decision variable of the problem. The target level strategy is based on making inventory for each echelon; in its turn order strategy is based on determination of optimal order quantity, which is independent from scenarios. However target level strategy provides high service at low cost, but it offers less reality under uncertain demand than order strategy. Practical experiments on finding the optimal SAA parameters are presented in the paper and as well as the analysis of their impact on solution quality.","PeriodicalId":211660,"journal":{"name":"Sci. J. Riga Tech. Univ. Ser. Comput. Sci.","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114259629","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 : 1900-01-01DOI: 10.2478/v10143-011-0029-y
Olegs Verhodubs, J. Grundspeņķis
Towards the Semantic Web Expert System The paper presents a conception of the Semantic Web Expert System which is the logical continuation of the expert system development. The Semantic Web Expert System emerges as the result of evolution of expert system concept and it means expert system moving toward the Web and using new Semantic Web technologies. The proposed conception of the Semantic Web Expert System promises to have new useful features that distinguish it from other types of expert systems. Semantiskā tīmekļa ekspertu sistēma Rakstā ir sniegts pārskats par ekspertu sistēmu attīstības tendencēm. Ir apskatītas tradicionālas un tīmekļa ekspertu sistēmas un raksta beigās ir piedāvātas semantiskā tīmekļa ekspertu sistēmas. Tradicionālās ir statiskās un dinamiskās ekspertu sistēmas, kuras strādā lokāli. Tīmekļa ekspertu sistēmas ir tādas sistēmas, kuras strādā tīmeklī. Semantiskā tīmekļa ekspertu sistēma ir ekspertu sistēma, kas strādā tīmeklī un lieto tādas tehnologijas, kā zināšanu atspoguļošana ar ontologiju aprakstīšanas valodu OWL un multiagentu sistēmu tehnologiju izmantošana. Rakstā ir izklāstītas aprakstīto ekspertu sistēmu struktūras, priekšrocības un trūkumi. Atsevišķi ir aprakstītas semantiskā tīmekļa tehnologijas, kuras ir piedāvāto semantiskā tīmekļa ekspertu sistēmu pamatā. Ir parādīts, kāpēc semantiskā tīmekļa tehnologijas var būt pielietotas ekspertu sistēmas konstruēšanai. Īpaša uzmanība ir pievērsta jaunajai ekspertu sistēmas koncepcijai un tās struktūrai, norādot sistēmas pamatkomponentes un to funkcijas. Rakstā detalizēti ir aprakstīti piedāvātās semantiskā tīmekļa ekspertu sistēmas funkcionēšanas režīmi. Ir norādītas semantiskā tīmekļa ekspertu sistēmas priekšrocības attiecībā pret tradicionālām un tīmekļa ekspertu sist ēmām, kā arī semantiskā tīmekļa ekspertu sistēmu konstruēšanas problēmas. Nobeigumā ir identificēti semantiskā tīmekļa ekspertu sistēmas turpmākās attīstības perspektīvas. Tiek domāts, ka piedāvātā semantiska tīmekļa ekspertu sistēma būs spējīga patstāvīgi papildināt savu zināšanas bāzi, lietojot tīmekļa resursus. Tas būs iespējams tikai tad, kad semantiska tīmekļa tehnologijas tiks plaši izmantotas tīmeklī. Экспертная система семантической сети В данной статье дан обзор тенденций развития экспертных систем, описаны обычные, сетевые и экспертные системы семантической сети. Под обычными экспертными системами понимаются статические и динамические экспертные системы, работающие локально. Под сетевыми понимаются экспертные системы, работающие в глобальной сети Интернет. Экспертная система семантической сети - это экспертная система, которая работает в сети Интернет и использует такие технологии как представление знаний с помощью языка описания онтологии OWL и использование мультиагентных систем. В этой статье изложены преимущества, недостатки и структуры описанных экспертных систем. Отдельно в статье описаны технологии семантической сети, которые лежат в основе предложенной экспертной системы семантической сети
{"title":"Towards the Semantic Web Expert System","authors":"Olegs Verhodubs, J. Grundspeņķis","doi":"10.2478/v10143-011-0029-y","DOIUrl":"https://doi.org/10.2478/v10143-011-0029-y","url":null,"abstract":"Towards the Semantic Web Expert System The paper presents a conception of the Semantic Web Expert System which is the logical continuation of the expert system development. The Semantic Web Expert System emerges as the result of evolution of expert system concept and it means expert system moving toward the Web and using new Semantic Web technologies. The proposed conception of the Semantic Web Expert System promises to have new useful features that distinguish it from other types of expert systems. Semantiskā tīmekļa ekspertu sistēma Rakstā ir sniegts pārskats par ekspertu sistēmu attīstības tendencēm. Ir apskatītas tradicionālas un tīmekļa ekspertu sistēmas un raksta beigās ir piedāvātas semantiskā tīmekļa ekspertu sistēmas. Tradicionālās ir statiskās un dinamiskās ekspertu sistēmas, kuras strādā lokāli. Tīmekļa ekspertu sistēmas ir tādas sistēmas, kuras strādā tīmeklī. Semantiskā tīmekļa ekspertu sistēma ir ekspertu sistēma, kas strādā tīmeklī un lieto tādas tehnologijas, kā zināšanu atspoguļošana ar ontologiju aprakstīšanas valodu OWL un multiagentu sistēmu tehnologiju izmantošana. Rakstā ir izklāstītas aprakstīto ekspertu sistēmu struktūras, priekšrocības un trūkumi. Atsevišķi ir aprakstītas semantiskā tīmekļa tehnologijas, kuras ir piedāvāto semantiskā tīmekļa ekspertu sistēmu pamatā. Ir parādīts, kāpēc semantiskā tīmekļa tehnologijas var būt pielietotas ekspertu sistēmas konstruēšanai. Īpaša uzmanība ir pievērsta jaunajai ekspertu sistēmas koncepcijai un tās struktūrai, norādot sistēmas pamatkomponentes un to funkcijas. Rakstā detalizēti ir aprakstīti piedāvātās semantiskā tīmekļa ekspertu sistēmas funkcionēšanas režīmi. Ir norādītas semantiskā tīmekļa ekspertu sistēmas priekšrocības attiecībā pret tradicionālām un tīmekļa ekspertu sist ēmām, kā arī semantiskā tīmekļa ekspertu sistēmu konstruēšanas problēmas. Nobeigumā ir identificēti semantiskā tīmekļa ekspertu sistēmas turpmākās attīstības perspektīvas. Tiek domāts, ka piedāvātā semantiska tīmekļa ekspertu sistēma būs spējīga patstāvīgi papildināt savu zināšanas bāzi, lietojot tīmekļa resursus. Tas būs iespējams tikai tad, kad semantiska tīmekļa tehnologijas tiks plaši izmantotas tīmeklī. Экспертная система семантической сети В данной статье дан обзор тенденций развития экспертных систем, описаны обычные, сетевые и экспертные системы семантической сети. Под обычными экспертными системами понимаются статические и динамические экспертные системы, работающие локально. Под сетевыми понимаются экспертные системы, работающие в глобальной сети Интернет. Экспертная система семантической сети - это экспертная система, которая работает в сети Интернет и использует такие технологии как представление знаний с помощью языка описания онтологии OWL и использование мультиагентных систем. В этой статье изложены преимущества, недостатки и структуры описанных экспертных систем. Отдельно в статье описаны технологии семантической сети, которые лежат в основе предложенной экспертной системы семантической сети","PeriodicalId":211660,"journal":{"name":"Sci. J. Riga Tech. Univ. Ser. Comput. Sci.","volume":"85 3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114991042","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 : 1900-01-01DOI: 10.2478/v10143-011-0045-y
P. Grabusts
Distance Metrics Selection Validity in Cluster Analysis In cluster analysis data are divided into groups according to a specific criterion called metrics. Traditionally the metrics of choice has been Euclidean distance. This article studies other distance metrics used in cluster analysis- Manhattan distance, Cosine distance and Pearson correlation measure. In k-means clustering algorithm these metrics were used to determine cluster centers and the clustering correctness was evaluated. It was found that the clustering results were very similar. The article also contemplates to evaluate clustering validity criteria. Attāluma metrikas izvēles pamatotība klasteranalīzē Klasteranalīzē ir nepieciešams kaut kādā veidā klasificēt datus vai atrast likumsakarības tajos, tāpēc jēdziens "likumsakarība" iegūst arvien lielāku nozīmi intelektuālās datu analīzes kontekstā. Bieži ir nepieciešams noskaidrot - kādā veidā dati ir saistīti savā starpā, kāda ir dažādu datu līdzība vai atšķirība, kāds ir šo datu salīdzināšanas mērs. Tādam nolūkam var izmantot dažādus klasterizācijas algoritmus, kas datus sadala grupās pēc noteiktiem kritērijiem - metrikas. Ar metriku šajā kontekstā tiek saprasta distance (attālums) starp klasterā ietilpstošajiem punktiem. Darbā tika pārbaudīta klasiskā klasterizācijas algoritma k-means darbības rezultāti ar dažādām metrikām: Eiklīda distanci, Manhattan distanci, Cosine distanci un Pīrsona korelācijas koeficientu. Eksperimentu gaitā k-means klasterizācijas algoritmā klasteru centru noteikšanai secīgi tika izmantotas minētās četras metrikas. Iegūtie rezultāti tika analizēti un tika pārbaudīts klasterizācijas korektums. Tradicionāli klasterizācijas algoritmos izmanto Eiklīda distanci, taču citas metrikas izvēle atsevišķos gadījumos var būt diskutējama. Tas atkarīgs no risināmā uzdevuma, datu apjoma un sarežgītības. Tika konstatēts, ka klasterizācijas rezultāti visu apskatāmo metriku izmantošanā ir ļoti līdzīgi. Nevienai no izvēlētajām metrikām nebija izšķirīga pārsvara, kas varētu garantēti pasludināt to par labāko. Darba izstrādes laikā aktualizējās jautājums par klasterizācijas kvalitātes kritērijiem, t.i., skaitliska kritērija noteikšanu, lai varētu novērtēt klasterizācijas rezultātu. Klasterizācijas kvalitātes kritēriji tika novērtēti ar Randa indeksu un Huberta indeksu. Обоснование выбора метрики расстояния в кластерном анализе В кластерном анализе необходимо каким-то образом классифицировать данные или найти в них закономерности, поэтому понятие закономерности имеет большое значение в контексте интеллектуальной обработки данных. Часто приходится выяснять каким образом данные связаны между собой, какова степень сходства или различия между ними, какова мера сравнения этих данных. Для таких целей можно использовать различные алгоритмы кластеризации, которые группируют данные по определенным критериям метрики. Под метрикой в этом контексте подразумевается расстояние (дистанция) между точками кластера. В статье проверяются результаты раб
{"title":"Distance Metrics Selection Validity in Cluster Analysis","authors":"P. Grabusts","doi":"10.2478/v10143-011-0045-y","DOIUrl":"https://doi.org/10.2478/v10143-011-0045-y","url":null,"abstract":"Distance Metrics Selection Validity in Cluster Analysis In cluster analysis data are divided into groups according to a specific criterion called metrics. Traditionally the metrics of choice has been Euclidean distance. This article studies other distance metrics used in cluster analysis- Manhattan distance, Cosine distance and Pearson correlation measure. In k-means clustering algorithm these metrics were used to determine cluster centers and the clustering correctness was evaluated. It was found that the clustering results were very similar. The article also contemplates to evaluate clustering validity criteria. Attāluma metrikas izvēles pamatotība klasteranalīzē Klasteranalīzē ir nepieciešams kaut kādā veidā klasificēt datus vai atrast likumsakarības tajos, tāpēc jēdziens \"likumsakarība\" iegūst arvien lielāku nozīmi intelektuālās datu analīzes kontekstā. Bieži ir nepieciešams noskaidrot - kādā veidā dati ir saistīti savā starpā, kāda ir dažādu datu līdzība vai atšķirība, kāds ir šo datu salīdzināšanas mērs. Tādam nolūkam var izmantot dažādus klasterizācijas algoritmus, kas datus sadala grupās pēc noteiktiem kritērijiem - metrikas. Ar metriku šajā kontekstā tiek saprasta distance (attālums) starp klasterā ietilpstošajiem punktiem. Darbā tika pārbaudīta klasiskā klasterizācijas algoritma k-means darbības rezultāti ar dažādām metrikām: Eiklīda distanci, Manhattan distanci, Cosine distanci un Pīrsona korelācijas koeficientu. Eksperimentu gaitā k-means klasterizācijas algoritmā klasteru centru noteikšanai secīgi tika izmantotas minētās četras metrikas. Iegūtie rezultāti tika analizēti un tika pārbaudīts klasterizācijas korektums. Tradicionāli klasterizācijas algoritmos izmanto Eiklīda distanci, taču citas metrikas izvēle atsevišķos gadījumos var būt diskutējama. Tas atkarīgs no risināmā uzdevuma, datu apjoma un sarežgītības. Tika konstatēts, ka klasterizācijas rezultāti visu apskatāmo metriku izmantošanā ir ļoti līdzīgi. Nevienai no izvēlētajām metrikām nebija izšķirīga pārsvara, kas varētu garantēti pasludināt to par labāko. Darba izstrādes laikā aktualizējās jautājums par klasterizācijas kvalitātes kritērijiem, t.i., skaitliska kritērija noteikšanu, lai varētu novērtēt klasterizācijas rezultātu. Klasterizācijas kvalitātes kritēriji tika novērtēti ar Randa indeksu un Huberta indeksu. Обоснование выбора метрики расстояния в кластерном анализе В кластерном анализе необходимо каким-то образом классифицировать данные или найти в них закономерности, поэтому понятие закономерности имеет большое значение в контексте интеллектуальной обработки данных. Часто приходится выяснять каким образом данные связаны между собой, какова степень сходства или различия между ними, какова мера сравнения этих данных. Для таких целей можно использовать различные алгоритмы кластеризации, которые группируют данные по определенным критериям метрики. Под метрикой в этом контексте подразумевается расстояние (дистанция) между точками кластера. В статье проверяются результаты раб","PeriodicalId":211660,"journal":{"name":"Sci. J. Riga Tech. Univ. Ser. Comput. Sci.","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115457849","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：