Multiple classifier systems (MCS) have become popular during the last decade. Self-generating neural tree (SGNT) is one of the suitable base-classifiers for MCS because of the simple setting and fast learning. In an e...
详细信息
ISBN:
(纸本)9781479959556
Multiple classifier systems (MCS) have become popular during the last decade. Self-generating neural tree (SGNT) is one of the suitable base-classifiers for MCS because of the simple setting and fast learning. In an earlier paper, we proposed a pruning method for the structure of the SGNT in the MCS to reduce the computational cost and we called this model as self-organizing neural grove (SONG). In this paper, we investigate a performance of incremental learning using SONG for two classification problems. The result shows that the SONG can reinsure rapid and efficient incremental learning.
A typical definitional question answering system extracts definition sentences from multiple documents and summarizes the sentences into definitions. In this paper, a new approach is proposed to extract definitional s...
详细信息
ISBN:
(纸本)9781424410651
A typical definitional question answering system extracts definition sentences from multiple documents and summarizes the sentences into definitions. In this paper, a new approach is proposed to extract definitional sentences from web corpus, which is based on softpattern and ontology. softpattern is used to judge whether sentences are definitional with syntactic analysis and ontology can deduce the definitional one with semantic information. The experimental result validates the effectiveness of this method, and the precision reaches 32.7% and is 5.2% higher than the method only based on softpattern.
This paper analyzes support vector machines (SVMs) and several commonly used softcomputing paradigms for patternrecognition including neural and wavelet networks, and fuzzy systems, Bayesian classifiers, fuzzy parti...
详细信息
ISBN:
(纸本)0819442836
This paper analyzes support vector machines (SVMs) and several commonly used softcomputing paradigms for patternrecognition including neural and wavelet networks, and fuzzy systems, Bayesian classifiers, fuzzy partitions, etc and tries to outline the similarities and differences among them. Support vector machines provide a new approach to the problem of patternrecognition with clear connections to the underlying statistical learning theory. We try to bring SVMs into the framework of the unification paradigm called the weighted radial basis function paradigm. Unifying different classes of methods has enormous advantages, such as the ability to merge all such techniques within the same system. It is hoped that this paper would provide theoretical guides for the study and applications of support vector machine and softcomputing paradigms.
This Volume 2 of the conference proceedings contains 110 papers. Topics discussed include fuzzy artificial intelligence, softcomputing applications to intelligent manufacturing and fault diagnosis, intelligent commun...
详细信息
This Volume 2 of the conference proceedings contains 110 papers. Topics discussed include fuzzy artificial intelligence, softcomputing applications to intelligent manufacturing and fault diagnosis, intelligent communication, fuzzy relations, patternrecognition, evolutionary algorithms, fuzzy modeling, math, softcomputing in medicine, control, finance markets, fuzzy set techniques in subjective evaluation and decision, neural networks and fuzzy techniques, neural networks and patternrecognition, fuzzy systems for intelligent control, information mining, softcomputing in image processing, fuzzy topology, genetic fuzzy systems and data mining and clustering.
In this paper, a review of softcomputing techniques in biometrics is presented. Biometrics has become one of the most promising authentication techniques in the last few years but issues like False Acceptance Rate, F...
详细信息
ISBN:
(纸本)9781467382533
In this paper, a review of softcomputing techniques in biometrics is presented. Biometrics has become one of the most promising authentication techniques in the last few years but issues like False Acceptance Rate, False Rejection Rate still prevails in biometrics. An efficient biometric system has higher recognition rate, tolerance for imprecision, uncertainty and noisy data. Recently, softcomputing has gained wide popularity in biometric recognition where it has helped in improving the recognition rate to a great extent. Various softcomputing techniques like fuzzy logic, evolutionary algorithm and artificial neural network has increasingly being used for the construction of efficient biometric systems. This paper first presents the introduction to biometrics along with the issues involved in it. A brief description of various softcomputing techniques for feature extraction, fusion, feature optimization, improvement of recognition rate in biometrics is provided. Finally future research areas are presented.
Pairwise dissimilarity representations are frequently used as an alternative to feature vectors in patternrecognition. One of the problems encountered in the analysis of such data, is that the dissimilarities are rar...
详细信息
ISBN:
(纸本)9783319023090
Pairwise dissimilarity representations are frequently used as an alternative to feature vectors in patternrecognition. One of the problems encountered in the analysis of such data, is that the dissimilarities are rarely Euclidean, while statistical learning algorithms often rely on Euclidean distances. Such non-Euclidean dissimilarities are often corrected or imposed geometry via embedding. This talk reviews and and extends the field of analysing non-Euclidean dissimilarity data.
We consider patternrecognition problem when classes and their labels are linearly structured (or ordered). We propose the loss function based on the squared differences between the true and the predicted class labels...
详细信息
ISBN:
(纸本)9783642132070
We consider patternrecognition problem when classes and their labels are linearly structured (or ordered). We propose the loss function based on the squared differences between the true and the predicted class labels. The optimal Bayes classifier is derived and then estimated by the recursive kernel estimator. Its consistency is established theoretically. Its RBF-like realization of the classifier is also proposed together with a recursive learning algorithm, which is well suited for on-line applications. The proposed approach was tested in real life example involving classification of moving vehicles.
Hidden Markov models are well-known methods for image processing. They are used in many areas where 1D data are processed. In the case of 2D data, there appear some problems with application HMM. There are some soluti...
详细信息
ISBN:
(纸本)9783642386572
Hidden Markov models are well-known methods for image processing. They are used in many areas where 1D data are processed. In the case of 2D data, there appear some problems with application HMM. There are some solutions, but they convert input observation from 2D to 1D, or create parallel pseudo 2D HMM, which is set of 1D HMMs in fact. This paper describes authentic 2D HMM with two-dimensional input data, and its application for patternrecognition in image processing.
In this paper, an walking pattern tuning system based on ZMP(Zero Moment Point) for a humanoid robot is proposed. Commonly, it is inevitable to handle the stage with proficient programming skill for realization of hum...
详细信息
ISBN:
(纸本)9781479959556
In this paper, an walking pattern tuning system based on ZMP(Zero Moment Point) for a humanoid robot is proposed. Commonly, it is inevitable to handle the stage with proficient programming skill for realization of humanoid walking. The proposed walking pattern tuning system is a program tool to finely tune the walking pattern based on ZMP by adjusting intuitive parameters that ordinary user can understand easily. The walking pattern tuning system was applied to the humanoid robot of 110cm tall and the humanoid robot with the tuned walking pattern was tested.
Purpose–This paper aims to consider a softcomputing approach to pattern classification using the basic tools of fuzzy relational calculus(FRC)and genetic algorithm(GA).Design/methodology/approach–The paper introduc...
详细信息
Purpose–This paper aims to consider a softcomputing approach to pattern classification using the basic tools of fuzzy relational calculus(FRC)and genetic algorithm(GA).Design/methodology/approach–The paper introduces a new interpretation of multidimensional fuzzy implication(MFI)to represent the author’s knowledge about the training data *** also considers the notion of a fuzzy pattern vector(FPV)to handle the fuzzy information granules of the quantized pattern space and to represent a population of training patterns in the quantized pattern *** construction of the pattern classifier is essentially based on the estimate of a fuzzy relation Ri between the antecedent clause and consequent clause of each one-dimensional fuzzy *** the estimation of Ri floating point representation of GA is ***,a set of fuzzy relations is formed from the new interpretation of *** set of fuzzy relations is termed as the core of the pattern *** the classifier is constructed the non-fuzzy features of a test pattern can be ***–The performance of the proposed scheme is tested on synthetic ***,the paper uses the proposed scheme for the vowel classification problem of an Indian *** all these case studies the recognition score of the proposed method is very ***,a benchmark of performance is established by considering Multilayer Perceptron(MLP),Support Vector Machine(SVM)and the proposed *** Abalone,Hosse colic and Pima Indians data sets,obtained from UCL database repository are used for the said benchmark *** benchmark study also establishes the superiority of the proposed ***/value–This new softcomputing approach to pattern classification is based on a new interpretation of MFI and a novel notion of FPV.A set of fuzzy relations which is the core of the pattern classifier,is estimated using floating point GA and very effective classification of patterns under vague and impreci
暂无评论