Pub Date : 2003-05-25DOI: 10.1109/FUZZ.2003.1206552
Dae-Jin Kim, Z. Bien, Kwang-Hyun Park
Facial expression recognition is very important in many human-robot/human-computer interaction systems. Although so many researches are done, it is hard to find a practical applications in the real world due to its underestimate about individual differences among people. Thus, as a solution for such problem, we introduce a 'personalized' facial expression recognition system. Many previous works on facial expression recognition focus on the well-known six universal facial expressions (happy, sad, fear, angry, surprise and disgust) under usage of unified (or non-separated) classification approach. However, for ordinary people, it is a very difficult task to make such facial expressions without much effort and training. Instead of universal facial expressions, many people show 'personalized' or 'individualized' facial expressions typically. Thus, for dealing with such personalities, we propose a method to construct a personalized classifier based on novel feature selection method. Specifically, feature selection is done by histogram-based approach in the frame of fuzzy neural networks(FNN). Besides, we also use an integrated approach for facial expression recognition. Actual experiments/simulations show that the proposed method is effective not only in view of facial expression recognition but also in view of pattern classifier itself.
{"title":"Fuzzy neural networks(FNN)-based approach for personalized facial expression recognition with novel feature selection method","authors":"Dae-Jin Kim, Z. Bien, Kwang-Hyun Park","doi":"10.1109/FUZZ.2003.1206552","DOIUrl":"https://doi.org/10.1109/FUZZ.2003.1206552","url":null,"abstract":"Facial expression recognition is very important in many human-robot/human-computer interaction systems. Although so many researches are done, it is hard to find a practical applications in the real world due to its underestimate about individual differences among people. Thus, as a solution for such problem, we introduce a 'personalized' facial expression recognition system. Many previous works on facial expression recognition focus on the well-known six universal facial expressions (happy, sad, fear, angry, surprise and disgust) under usage of unified (or non-separated) classification approach. However, for ordinary people, it is a very difficult task to make such facial expressions without much effort and training. Instead of universal facial expressions, many people show 'personalized' or 'individualized' facial expressions typically. Thus, for dealing with such personalities, we propose a method to construct a personalized classifier based on novel feature selection method. Specifically, feature selection is done by histogram-based approach in the frame of fuzzy neural networks(FNN). Besides, we also use an integrated approach for facial expression recognition. Actual experiments/simulations show that the proposed method is effective not only in view of facial expression recognition but also in view of pattern classifier itself.","PeriodicalId":212172,"journal":{"name":"The 12th IEEE International Conference on Fuzzy Systems, 2003. FUZZ '03.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2003-05-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129829421","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 : 2003-05-25DOI: 10.1109/FUZZ.2003.1206655
O. Castillo, P. Melin
This tutorial will show how to use different Soil Computing (SC) techniques for the development of hybrid intelligent systems for industrial applications. SC techniques, at the moment, include Neural Networks, Fuzzy Logic, Genetic Algorithms and Chaos Theory. We also consider the use of Fractal Theory for pattern recognition and time series analysis. Each of these methodologies has its advantages and disadvantages and many problems have been solved, by using one of these methodologies. However, many real-world complex industrial problems require the integration of several of these methodologies to really achieve the efficiency and accuracy needed in practice. In this tutorial a brief introduction to SC methodologies will be given. Then, different methods for integrating the different SC methodologies in solving real-world problems will be described. At the end, the integration methodologies will he illustrated with real hybrid intelligent systems that have been developed for applications like: Food Processing Plants, Robotic Systems, Automated Quality Control, Financial and Economic Forecasting, and Manufacturing Systems, Those attending can expect to gain awareness of the role of SC methodologies and their integration in solving real world complex problems.
{"title":"Soft computing and fractal theory for industrial applications","authors":"O. Castillo, P. Melin","doi":"10.1109/FUZZ.2003.1206655","DOIUrl":"https://doi.org/10.1109/FUZZ.2003.1206655","url":null,"abstract":"This tutorial will show how to use different Soil Computing (SC) techniques for the development of hybrid intelligent systems for industrial applications. SC techniques, at the moment, include Neural Networks, Fuzzy Logic, Genetic Algorithms and Chaos Theory. We also consider the use of Fractal Theory for pattern recognition and time series analysis. Each of these methodologies has its advantages and disadvantages and many problems have been solved, by using one of these methodologies. However, many real-world complex industrial problems require the integration of several of these methodologies to really achieve the efficiency and accuracy needed in practice. In this tutorial a brief introduction to SC methodologies will be given. Then, different methods for integrating the different SC methodologies in solving real-world problems will be described. At the end, the integration methodologies will he illustrated with real hybrid intelligent systems that have been developed for applications like: Food Processing Plants, Robotic Systems, Automated Quality Control, Financial and Economic Forecasting, and Manufacturing Systems, Those attending can expect to gain awareness of the role of SC methodologies and their integration in solving real world complex problems.","PeriodicalId":212172,"journal":{"name":"The 12th IEEE International Conference on Fuzzy Systems, 2003. FUZZ '03.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2003-05-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127155402","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 : 2003-05-25DOI: 10.1109/FUZZ.2003.1209395
R. Palm
Optimization and control of systems is a challenge in the case of a large number of complex local systems. Decentralized methods like multi-agent control are expected to handle optimization tasks more efficiently than centralized approaches. One of the most interesting decentralized methods is the market-based optimization. Market-based algorithms imitate the behavior of economic systems in which virtual producer and consumer agents compete and cooperate. This method is applied to the optimization and synchronization of a set of local multiple-model systems and Takagi-Sugeno (TS) fuzzy subsystems. Implementing a ring structure between the local systems leads to the same results as for the unrestricted net structure.
{"title":"Optimization of decentralized multiple-model systems and TS fuzzy systems","authors":"R. Palm","doi":"10.1109/FUZZ.2003.1209395","DOIUrl":"https://doi.org/10.1109/FUZZ.2003.1209395","url":null,"abstract":"Optimization and control of systems is a challenge in the case of a large number of complex local systems. Decentralized methods like multi-agent control are expected to handle optimization tasks more efficiently than centralized approaches. One of the most interesting decentralized methods is the market-based optimization. Market-based algorithms imitate the behavior of economic systems in which virtual producer and consumer agents compete and cooperate. This method is applied to the optimization and synchronization of a set of local multiple-model systems and Takagi-Sugeno (TS) fuzzy subsystems. Implementing a ring structure between the local systems leads to the same results as for the unrestricted net structure.","PeriodicalId":212172,"journal":{"name":"The 12th IEEE International Conference on Fuzzy Systems, 2003. FUZZ '03.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2003-05-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127220976","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 : 2003-05-25DOI: 10.1109/FUZZ.2003.1206539
A. Soria-Frisch
The construction of fuzzy measures in the fuzzy integral, which is considered to be the crucial point for the extended utilization of this fusion methodology, is attained in the here presented paper through a Self-Organizing Map (SOM). This fact can improve the performance in the fuzzy measure assessment specially in high-dimensional feature spaces. Different methodologies for knowledge discovery related to the SOM paradigm are taken into consideration in order to achieve the assessment of the fuzzy measure coefficients. Furthermore an overview of the utilization of the fuzzy integral in classification problems is given. Finally two hybrid frameworks considering the SOM and the fuzzy integral are presented and used for fuzzy classification.
{"title":"Hybrid SOM and fuzzy integral frameworks for fuzzy classification","authors":"A. Soria-Frisch","doi":"10.1109/FUZZ.2003.1206539","DOIUrl":"https://doi.org/10.1109/FUZZ.2003.1206539","url":null,"abstract":"The construction of fuzzy measures in the fuzzy integral, which is considered to be the crucial point for the extended utilization of this fusion methodology, is attained in the here presented paper through a Self-Organizing Map (SOM). This fact can improve the performance in the fuzzy measure assessment specially in high-dimensional feature spaces. Different methodologies for knowledge discovery related to the SOM paradigm are taken into consideration in order to achieve the assessment of the fuzzy measure coefficients. Furthermore an overview of the utilization of the fuzzy integral in classification problems is given. Finally two hybrid frameworks considering the SOM and the fuzzy integral are presented and used for fuzzy classification.","PeriodicalId":212172,"journal":{"name":"The 12th IEEE International Conference on Fuzzy Systems, 2003. FUZZ '03.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2003-05-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132149811","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 : 2003-05-25DOI: 10.1109/FUZZ.2003.1206547
Mehmet Kaya, R. Alhajj
In this paper, we propose genetic algorithms (GAs) based clustering method, which dynamically adjusts the fuzzy sets to provide maximum profit within an interval of user specified minimum support values. This is achieved by tuning the base values of the membership functions for each quantitative attribute so as to maximize the sum of large itemsets in a certain interval of minimum support values. To the best of our knowledge, this is the first effort in this direction. To support our claim, we compare the proposed GAs-based approach with a CURE-based approach. Experimental results on synthetic transactions show that the proposed clustering method exhibits a good performance over CURE-based approach in terms of the number of produced large itemsets and interesting association rules.
{"title":"A clustering algorithm with genetically optimized membership functions for fuzzy association rules mining","authors":"Mehmet Kaya, R. Alhajj","doi":"10.1109/FUZZ.2003.1206547","DOIUrl":"https://doi.org/10.1109/FUZZ.2003.1206547","url":null,"abstract":"In this paper, we propose genetic algorithms (GAs) based clustering method, which dynamically adjusts the fuzzy sets to provide maximum profit within an interval of user specified minimum support values. This is achieved by tuning the base values of the membership functions for each quantitative attribute so as to maximize the sum of large itemsets in a certain interval of minimum support values. To the best of our knowledge, this is the first effort in this direction. To support our claim, we compare the proposed GAs-based approach with a CURE-based approach. Experimental results on synthetic transactions show that the proposed clustering method exhibits a good performance over CURE-based approach in terms of the number of produced large itemsets and interesting association rules.","PeriodicalId":212172,"journal":{"name":"The 12th IEEE International Conference on Fuzzy Systems, 2003. FUZZ '03.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2003-05-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127946401","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 : 2003-05-25DOI: 10.1109/FUZZ.2003.1209392
Kaoping Song, Jicheng Zhang, Erlong Yang
Fracturing is an effective way to enhance oil recovery and widely used in the oil field. A good fracturing result relies on proper selection of target well and target layer. Traditionally, the method to selecting target wells and target layers is always done artificially and imprecisely. As a result, fracturing does not perform as well as it can. This paper proposes a fuzzy method of optimizing the selection of target wells and target layers for fracturing. This study successfully applies the fuzzy theory to the job to fracturing. It is an important technique of fracturing.
{"title":"A fuzzy optimization model to select the fracturing layers","authors":"Kaoping Song, Jicheng Zhang, Erlong Yang","doi":"10.1109/FUZZ.2003.1209392","DOIUrl":"https://doi.org/10.1109/FUZZ.2003.1209392","url":null,"abstract":"Fracturing is an effective way to enhance oil recovery and widely used in the oil field. A good fracturing result relies on proper selection of target well and target layer. Traditionally, the method to selecting target wells and target layers is always done artificially and imprecisely. As a result, fracturing does not perform as well as it can. This paper proposes a fuzzy method of optimizing the selection of target wells and target layers for fracturing. This study successfully applies the fuzzy theory to the job to fracturing. It is an important technique of fracturing.","PeriodicalId":212172,"journal":{"name":"The 12th IEEE International Conference on Fuzzy Systems, 2003. FUZZ '03.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2003-05-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128773725","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 : 2003-05-25DOI: 10.1109/FUZZ.2003.1206542
H. Frigui, S.A. Salem
In this paper, we propose an algorithm that performs fuzzy clustering and feature weighting simultaneously and in an unsupervised manner. The feature set is divided into logical subsets of features, and a degree of relevance is dynamically assigned to each subset based on its partial degree of dissimilarity. The proposed algorithm is computationally and implementationally simple, and learns a different set of feature weights for each cluster. The cluster dependent feature weights have two advantages. First, they help in partitioning the data set into more meaningful clusters. Second, they can be used as part of a more complex learning system to enhance its learning behavior. The performance of the proposed algorithm is illustrated by using it to segment color images.
{"title":"Fuzzy clustering and subset feature weighting","authors":"H. Frigui, S.A. Salem","doi":"10.1109/FUZZ.2003.1206542","DOIUrl":"https://doi.org/10.1109/FUZZ.2003.1206542","url":null,"abstract":"In this paper, we propose an algorithm that performs fuzzy clustering and feature weighting simultaneously and in an unsupervised manner. The feature set is divided into logical subsets of features, and a degree of relevance is dynamically assigned to each subset based on its partial degree of dissimilarity. The proposed algorithm is computationally and implementationally simple, and learns a different set of feature weights for each cluster. The cluster dependent feature weights have two advantages. First, they help in partitioning the data set into more meaningful clusters. Second, they can be used as part of a more complex learning system to enhance its learning behavior. The performance of the proposed algorithm is illustrated by using it to segment color images.","PeriodicalId":212172,"journal":{"name":"The 12th IEEE International Conference on Fuzzy Systems, 2003. FUZZ '03.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2003-05-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126582814","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 : 2003-05-25DOI: 10.1109/FUZZ.2003.1209409
J. M. Adams, K. Rattan
A multi-stage fuzzy logic model is systematically reduced to obtain a significantly smaller rulebase. The multi-stage structure is obtained by unfolding a single-stage, n-dimension fuzzy logic model into multiple, two-dimension stages. The interconnection between stages is not defuzzified. Rule reduction is performed by comparing output membership functions in the final two-dimension rulebase, weighted by the amount of use each rule receives, called the sum of truth, from the previous stage. The method is demonstrated on a Mackey Glass series and on a two-link robot, both with encouraging results.
{"title":"Systematic rule reduction of a multi-stage fuzzy logic model","authors":"J. M. Adams, K. Rattan","doi":"10.1109/FUZZ.2003.1209409","DOIUrl":"https://doi.org/10.1109/FUZZ.2003.1209409","url":null,"abstract":"A multi-stage fuzzy logic model is systematically reduced to obtain a significantly smaller rulebase. The multi-stage structure is obtained by unfolding a single-stage, n-dimension fuzzy logic model into multiple, two-dimension stages. The interconnection between stages is not defuzzified. Rule reduction is performed by comparing output membership functions in the final two-dimension rulebase, weighted by the amount of use each rule receives, called the sum of truth, from the previous stage. The method is demonstrated on a Mackey Glass series and on a two-link robot, both with encouraging results.","PeriodicalId":212172,"journal":{"name":"The 12th IEEE International Conference on Fuzzy Systems, 2003. FUZZ '03.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2003-05-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122248006","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 : 2003-05-25DOI: 10.1109/FUZZ.2003.1206565
F. Rhee, Yong-Shik Shin
In this paper, we propose a fast numerical algorithm for finding the optimal threshold for segmenting gray scale images. In the proposed method, several fuzzy entropy measures are introduced and the objective is to locate the gray level that possesses the minimum entropy. Instead of having to calculate the entropy for every gray level and determining the gray level where the entropy is minimum, the fixed point iteration (FPI) method is used to significantly speed up the process. In doing so, the optimal threshold may be quickly obtained within a few number of evaluations. To show the validity of our proposed algorithm, we test 7 types of fuzzy entropy measures on several images. The experimental results show that the proposed algorithm is much faster without loss of performance than the methods in earlier surveys.
{"title":"A fast numerical method for finding the optimal threshold for image segmentation","authors":"F. Rhee, Yong-Shik Shin","doi":"10.1109/FUZZ.2003.1206565","DOIUrl":"https://doi.org/10.1109/FUZZ.2003.1206565","url":null,"abstract":"In this paper, we propose a fast numerical algorithm for finding the optimal threshold for segmenting gray scale images. In the proposed method, several fuzzy entropy measures are introduced and the objective is to locate the gray level that possesses the minimum entropy. Instead of having to calculate the entropy for every gray level and determining the gray level where the entropy is minimum, the fixed point iteration (FPI) method is used to significantly speed up the process. In doing so, the optimal threshold may be quickly obtained within a few number of evaluations. To show the validity of our proposed algorithm, we test 7 types of fuzzy entropy measures on several images. The experimental results show that the proposed algorithm is much faster without loss of performance than the methods in earlier surveys.","PeriodicalId":212172,"journal":{"name":"The 12th IEEE International Conference on Fuzzy Systems, 2003. FUZZ '03.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2003-05-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121413775","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 : 2003-05-25DOI: 10.1109/FUZZ.2003.1209315
A. Mamdani
Summary form only given. The occasion of this presentation provides an opportunity to re visit the first application of fuzzy control to a model steam engine carried out back in 1972 - 30 years ago. That work was focused on a problem that was looking for a solution which ultimately resulted in the use of fuzzy rules. The presentation will describe how the rules can be seen as an embodiment of the ownership of the problem. The solution (FLC) came to be a technique in its own right that has been applied to a vast number of other problems. FLC as with many other novel applications has had its share of serendipity. Beyond the reminiscences, the presentation will try and look for some lessons that can be learnt. The theoretically inclined (who provide the solution designers with the mathematical help needed) and the problem owners, even while dealing with the same technique, have different aims and expectations and indeed inhabit different worlds. Yet they can come together to face the challenges of new problems and provide solutions which benefit us all.
{"title":"Fuzzy logic control - from owning the problem to finding a good solution","authors":"A. Mamdani","doi":"10.1109/FUZZ.2003.1209315","DOIUrl":"https://doi.org/10.1109/FUZZ.2003.1209315","url":null,"abstract":"Summary form only given. The occasion of this presentation provides an opportunity to re visit the first application of fuzzy control to a model steam engine carried out back in 1972 - 30 years ago. That work was focused on a problem that was looking for a solution which ultimately resulted in the use of fuzzy rules. The presentation will describe how the rules can be seen as an embodiment of the ownership of the problem. The solution (FLC) came to be a technique in its own right that has been applied to a vast number of other problems. FLC as with many other novel applications has had its share of serendipity. Beyond the reminiscences, the presentation will try and look for some lessons that can be learnt. The theoretically inclined (who provide the solution designers with the mathematical help needed) and the problem owners, even while dealing with the same technique, have different aims and expectations and indeed inhabit different worlds. Yet they can come together to face the challenges of new problems and provide solutions which benefit us all.","PeriodicalId":212172,"journal":{"name":"The 12th IEEE International Conference on Fuzzy Systems, 2003. FUZZ '03.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2003-05-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121498890","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：