Pub Date : 2008-03-04DOI: 10.1109/GEFS.2008.4484565
P. Angelov, A. Kordon, Xiaowei Zhou
This paper describes an approach to design self-developing and self-tuning inferential soft sensors applicable to process industries. The proposal is for a Takagi-Sugeno-fuzzy system framework that has evolving (open structure) architecture, and an on-line (possibly real-time) learning algorithm. The proposed methodology is novel and it addresses the problems of self-development and self-calibration caused by drift in the data patterns due to changes in the operating regimes, catalysts ageing, industrial equipment wearing, contamination etc. The proposed computational technique is data-driven and parameter-free (it only requires a couple of parameters with clear meaning and suggested values). In this paper a case study of four problems of estimation of chemical properties is considered, however, the methodology has a much wider validity. The optimal inputs to the proposed evolving inferential sensor are determined a priori and off-line using a multi-objective genetic-programming-based optimization. Different on-line input selection techniques are under development. The methodology is validated on real data provided by the Dow Chemical Company, USA.
{"title":"Evolving fuzzy inferential sensors for process industry","authors":"P. Angelov, A. Kordon, Xiaowei Zhou","doi":"10.1109/GEFS.2008.4484565","DOIUrl":"https://doi.org/10.1109/GEFS.2008.4484565","url":null,"abstract":"This paper describes an approach to design self-developing and self-tuning inferential soft sensors applicable to process industries. The proposal is for a Takagi-Sugeno-fuzzy system framework that has evolving (open structure) architecture, and an on-line (possibly real-time) learning algorithm. The proposed methodology is novel and it addresses the problems of self-development and self-calibration caused by drift in the data patterns due to changes in the operating regimes, catalysts ageing, industrial equipment wearing, contamination etc. The proposed computational technique is data-driven and parameter-free (it only requires a couple of parameters with clear meaning and suggested values). In this paper a case study of four problems of estimation of chemical properties is considered, however, the methodology has a much wider validity. The optimal inputs to the proposed evolving inferential sensor are determined a priori and off-line using a multi-objective genetic-programming-based optimization. Different on-line input selection techniques are under development. The methodology is validated on real data provided by the Dow Chemical Company, USA.","PeriodicalId":343300,"journal":{"name":"2008 3rd International Workshop on Genetic and Evolving Systems","volume":"128 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117070206","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 : 2008-03-04DOI: 10.1109/GEFS.2008.4484574
M. Mucientes, J. Vidal, Alberto Bugarín-Diz, M. Lama
Accuracy in processing time estimation of manufacturing operations is fundamental to achieve more competitive prices and higher profits in an industry. The manufacturing times of a machine depend on several input variables and, for each class or type of product, a regression function for that machine can be defined. Time estimations are used for implementing production plans. These plans are usually supervised and modified by an expert, so information about the dependencies of processing time with the input variables is also very important. Taking into account both premises (accuracy and simplicity in information extraction), a model based on TSK (Takagi-Sugeno-Kang) fuzzy rules has been used. TSK rules fulfill both requisites: the system has a high accuracy, and the knowledge structure makes explicit the dependencies between time estimations and the input variables. We propose a TSK fuzzy rule model in which the rules have a variable structure in the consequent, as the regression functions can be completely distinct for different machines or, even, for different classes of inputs to the same machine. The methodology to learn the TSK knowledge base is based on genetic programming together with a context-free grammar to restrict the valid structures of the regression functions. The system has been tested with real data coming from five different machines of a wood furniture industry.
{"title":"Processing times estimation in a manufacturing industry through genetic programming","authors":"M. Mucientes, J. Vidal, Alberto Bugarín-Diz, M. Lama","doi":"10.1109/GEFS.2008.4484574","DOIUrl":"https://doi.org/10.1109/GEFS.2008.4484574","url":null,"abstract":"Accuracy in processing time estimation of manufacturing operations is fundamental to achieve more competitive prices and higher profits in an industry. The manufacturing times of a machine depend on several input variables and, for each class or type of product, a regression function for that machine can be defined. Time estimations are used for implementing production plans. These plans are usually supervised and modified by an expert, so information about the dependencies of processing time with the input variables is also very important. Taking into account both premises (accuracy and simplicity in information extraction), a model based on TSK (Takagi-Sugeno-Kang) fuzzy rules has been used. TSK rules fulfill both requisites: the system has a high accuracy, and the knowledge structure makes explicit the dependencies between time estimations and the input variables. We propose a TSK fuzzy rule model in which the rules have a variable structure in the consequent, as the regression functions can be completely distinct for different machines or, even, for different classes of inputs to the same machine. The methodology to learn the TSK knowledge base is based on genetic programming together with a context-free grammar to restrict the valid structures of the regression functions. The system has been tested with real data coming from five different machines of a wood furniture industry.","PeriodicalId":343300,"journal":{"name":"2008 3rd International Workshop on Genetic and Evolving Systems","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114507648","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 : 2008-03-04DOI: 10.1109/GEFS.2008.4484575
F. Berlanga, M. J. Jesús, F. Herrera
In this contribution, we present GP-COACH, a novel GFS based on the cooperative-competitive learning approach, that uses genetic programming to code fuzzy rules with a different number of variables, for getting compact and accurate rule bases for high dimensional problems. GP-COACH learns disjunctive normal form rules (generated by means of a context-free grammar) and uses a token competition mechanism to maintain the diversity of the population. It makes the rules compete and cooperate among themselves, giving out a compact set of fuzzy rules that presents a good performance. The good results obtained in an experimental study involving several high dimensional classification problems support our proposal.
{"title":"A novel genetic cooperative-competitive fuzzy rule based learning method using genetic programming for high dimensional problems","authors":"F. Berlanga, M. J. Jesús, F. Herrera","doi":"10.1109/GEFS.2008.4484575","DOIUrl":"https://doi.org/10.1109/GEFS.2008.4484575","url":null,"abstract":"In this contribution, we present GP-COACH, a novel GFS based on the cooperative-competitive learning approach, that uses genetic programming to code fuzzy rules with a different number of variables, for getting compact and accurate rule bases for high dimensional problems. GP-COACH learns disjunctive normal form rules (generated by means of a context-free grammar) and uses a token competition mechanism to maintain the diversity of the population. It makes the rules compete and cooperate among themselves, giving out a compact set of fuzzy rules that presents a good performance. The good results obtained in an experimental study involving several high dimensional classification problems support our proposal.","PeriodicalId":343300,"journal":{"name":"2008 3rd International Workshop on Genetic and Evolving Systems","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123503905","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 : 2008-03-04DOI: 10.1109/GEFS.2008.4484560
O. Cordón, A. Quirin, L. Sánchez
Fuzzy rule-based classification systems (FRBCSs) are able to design interpretable classifiers but suffer from the curse of dimensionality when dealing with complex problems with a large number of features. In this contribution we explore the use of popular approaches for designing ensembles of classifiers in the machine learning field, bagging and random subspace, to design FRBCS multiclassifiers from a basic, heuristic fuzzy classification rule generation method, aiming to both improve their accuracy and to make them able to deal with high dimensional classification problems. Besides, a multicriteria genetic algorithm is proposed to select the component classifiers in the ensemble guided by the cumulative likelihood in order to look for an appropriate accuracy-complexity trade-off.
{"title":"A first study on bagging fuzzy rule-based classification systems with multicriteria genetic selection of the component classifiers","authors":"O. Cordón, A. Quirin, L. Sánchez","doi":"10.1109/GEFS.2008.4484560","DOIUrl":"https://doi.org/10.1109/GEFS.2008.4484560","url":null,"abstract":"Fuzzy rule-based classification systems (FRBCSs) are able to design interpretable classifiers but suffer from the curse of dimensionality when dealing with complex problems with a large number of features. In this contribution we explore the use of popular approaches for designing ensembles of classifiers in the machine learning field, bagging and random subspace, to design FRBCS multiclassifiers from a basic, heuristic fuzzy classification rule generation method, aiming to both improve their accuracy and to make them able to deal with high dimensional classification problems. Besides, a multicriteria genetic algorithm is proposed to select the component classifiers in the ensemble guided by the cumulative likelihood in order to look for an appropriate accuracy-complexity trade-off.","PeriodicalId":343300,"journal":{"name":"2008 3rd International Workshop on Genetic and Evolving Systems","volume":"174 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130883519","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 : 2008-03-04DOI: 10.1109/GEFS.2008.4484570
T. Lasota, B. Trawinski, Krzysztof Trawiński
Takagi-Sugeno-Kang-type fuzzy model to assist with real estate appraisals is described and optimized using evolutionary algorithms Two approaches were compared in the paper. The first one consisted in learning the rule base and the second one in combining learning the rule base and tuning the membership functions in one process. Five TSK-type fuzzy models comprising 3 or 4 input variables referring to the attributes of a property were evaluated. The evolutionary algorithms were based on Pittsburgh approach with the real coded chromosomes of constant length comprising whole rule base or both the rule base and all parameters of all membership functions. The experiments were conducted using training and testing sets prepared on the basis of actual 134 sales transactions made in one of Polish cities and located in a residential section.
{"title":"Evolutionary generation of rule base in TSK fuzzy model for real estate appraisal","authors":"T. Lasota, B. Trawinski, Krzysztof Trawiński","doi":"10.1109/GEFS.2008.4484570","DOIUrl":"https://doi.org/10.1109/GEFS.2008.4484570","url":null,"abstract":"Takagi-Sugeno-Kang-type fuzzy model to assist with real estate appraisals is described and optimized using evolutionary algorithms Two approaches were compared in the paper. The first one consisted in learning the rule base and the second one in combining learning the rule base and tuning the membership functions in one process. Five TSK-type fuzzy models comprising 3 or 4 input variables referring to the attributes of a property were evaluated. The evolutionary algorithms were based on Pittsburgh approach with the real coded chromosomes of constant length comprising whole rule base or both the rule base and all parameters of all membership functions. The experiments were conducted using training and testing sets prepared on the basis of actual 134 sales transactions made in one of Polish cities and located in a residential section.","PeriodicalId":343300,"journal":{"name":"2008 3rd International Workshop on Genetic and Evolving Systems","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114691104","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 : 2008-03-04DOI: 10.1109/GEFS.2008.4484572
J. Alcalá-Fdez, S. García, F. Berlanga, Alberto Fernández, L. Sánchez, M. J. Jesús, F. Herrera
This work introduces the software tool KEEL to assess evolutionary algorithms for data mining problems including regression, classification, clustering, pattern mining and so on. It includes a big collection of genetic fuzzy system algorithms based on different approaches: Pittsburgh, Michigan, IRL and GCCL. It allows us to perform a complete analysis of any genetic fuzzy system in comparison to existing ones, including a statistical test module for comparison. The use of KEEL is illustrated through the analysis of one case study.
{"title":"KEEL: A data mining software tool integrating genetic fuzzy systems","authors":"J. Alcalá-Fdez, S. García, F. Berlanga, Alberto Fernández, L. Sánchez, M. J. Jesús, F. Herrera","doi":"10.1109/GEFS.2008.4484572","DOIUrl":"https://doi.org/10.1109/GEFS.2008.4484572","url":null,"abstract":"This work introduces the software tool KEEL to assess evolutionary algorithms for data mining problems including regression, classification, clustering, pattern mining and so on. It includes a big collection of genetic fuzzy system algorithms based on different approaches: Pittsburgh, Michigan, IRL and GCCL. It allows us to perform a complete analysis of any genetic fuzzy system in comparison to existing ones, including a statistical test module for comparison. The use of KEEL is illustrated through the analysis of one case study.","PeriodicalId":343300,"journal":{"name":"2008 3rd International Workshop on Genetic and Evolving Systems","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121660074","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：