• 한국데이터정보과학회:학술대회논문집
• /
• 2005.04a
• /
• pp.83-126
• /
• 2005

Single-index models have found applications in econometrics and biometrics, where multidimensional regression models are often encountered. Here we propose a nonparametric estimation approach that combines wavelet methods for non-equispaced designs with Bayesian models. We consider a wavelet series expansion of the unknown regression function and set prior distributions for the wavelet coefficients and the other model parameters. To ensure model identifiability, the direction parameter is represented via its polar coordinates. We employ ad hoc hierarchical mixture priors that perform shrinkage on wavelet coefficients and use Markov chain Monte Carlo methods for a posteriori inference. We investigate an independence-type Metropolis-Hastings algorithm to produce samples for the direction parameter. Our method leads to simultaneous estimates of the link function and of the index parameters. We present results on both simulated and real data, where we look at comparisons with other methods.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.22 no.4
• /
• pp.399-404
• /
• 2012

Evolutionary algorithms such as genetic algorithm (GA) have been proven their effectiveness when applying to the design of fuzzy models. However, it tends to suffer from computationally expensWive due to the slow convergence speed. In this study, we propose an approach to develop fuzzy models by means of an improved differential evolution (IDE) to overcome this limitation. The improved differential evolution (IDE) is realized by means of an orthogonal approach and differential evolution. With the invoking orthogonal method, the IDE can search the solution space more efficiently. In the design of fuzzy models, we concern two mechanisms, namely structure identification and parameter estimation. The structure identification is supported by the IDE and C-Means while the parameter estimation is realized via IDE and a standard least square error method. Experimental studies demonstrate that the proposed model leads to improved performance. The proposed model is also contrasted with the quality of some fuzzy models already reported in the literature.

• Journal of the Korean Data and Information Science Society
• /
• v.12 no.1
• /
• pp.19-25
• /
• 2001

Random effects models which describe the dependence via random effects in various correlated data have recently received considerable attention in the biomedical literature. They include mixed linear models (MLMs), generatized linear mixed models (GLMMS) and hierarchical generalized linear models (HGLMs). For the inference Lee and Nelder (2000) proposed the first-and second-order REML (restricted maximum likelihood) methods based on hierarchical-likelihood of tee and Welder (1996). In this paper, for Poisson-gamma HGLMs the new methods are theoretically compared with marginal likelihood methods and both methods are illustrated by two practical examples.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.21 no.6
• /
• pp.686-691
• /
• 2011

This study is concerned with the identification of fuzzy models. To address the optimization of fuzzy model, we proposed an improved space search evolutionary algorithm (ISSA) which is realized with the combination of space search algorithm and Gaussian mutation. The proposed ISSA is exploited here as the optimization vehicle for the design of fuzzy models. Considering the design of fuzzy models, we developed a hybrid identification method using information granulation and the ISSA. Information granules are treated as collections of objects (e.g. data) brought together by the criteria of proximity, similarity, or functionality. The overall hybrid identification comes in the form of two optimization mechanisms: structure identification and parameter identification. The structure identification is supported by the ISSA and C-Means while the parameter estimation is realized via the ISSA and weighted least square error method. A suite of comparative studies show that the proposed model leads to better performance in comparison with some existing models.

• Journal of Korea Multimedia Society
• /
• v.11 no.6
• /
• pp.807-815
• /
• 2008

A simple efficient method of spotting and recognizing hand gestures in video is presented using a network of hidden Markov models and dynamic programming search algorithm. The description starts from designing a set of isolated trajectory models which are stochastic and robust enough to characterize highly variable patterns like human motion, handwriting, and speech. Those models are interconnected to form a single big network termed a spotting network or a spotter that models a continuous stream of gestures and non-gestures as well. The inference over the model is based on dynamic programming. The proposed model is highly efficient and can readily be extended to a variety of recurrent pattern recognition tasks. The test result without any engineering has shown the potential for practical application. At the end of the paper we add some related experimental result that has been obtained using a different model - dynamic Bayesian network - which is also a type of stochastic model.

• Communications for Statistical Applications and Methods
• /
• v.29 no.5
• /
• pp.513-532
• /
• 2022

In this paper, the use of Bayes factors and the deviance information criterion for model selection are compared in a Bayesian accelerated life testing setup. In Bayesian accelerated life testing, the most used tool for model comparison is the deviance information criterion. An alternative and more formal approach is to use Bayes factors to compare models. However, Bayesian accelerated life testing models with more than one stressor often have mathematically intractable posterior distributions and Markov chain Monte Carlo methods are employed to obtain posterior samples to base inference on. The computation of the marginal likelihood is challenging when working with such complex models. In this paper, methods for approximating the marginal likelihood and the application thereof in the accelerated life testing paradigm are explored for dual-stress models. A simulation study is also included, where Bayes factors using the different approximation methods and the deviance information are compared.

• Journal of Electrical Engineering and Technology
• /
• v.6 no.6
• /
• pp.853-866
• /
• 2011

We propose a multi-objective space search algorithm (MSSA) and introduce the identification of fuzzy inference systems based on the MSSA and information granulation (IG). The MSSA is a multi-objective optimization algorithm whose search method is associated with the analysis of the solution space. The multi-objective mechanism of MSSA is realized using a non-dominated sorting-based multi-objective strategy. In the identification of the fuzzy inference system, the MSSA is exploited to carry out parametric optimization of the fuzzy model and to achieve its structural optimization. The granulation of information is attained using the C-Means clustering algorithm. The overall optimization of fuzzy inference systems comes in the form of two identification mechanisms: structure identification (such as the number of input variables to be used, a specific subset of input variables, the number of membership functions, and the polynomial type) and parameter identification (viz. the apexes of membership function). The structure identification is developed by the MSSA and C-Means, whereas the parameter identification is realized via the MSSA and least squares method. The evaluation of the performance of the proposed model was conducted using three representative numerical examples such as gas furnace, NOx emission process data, and Mackey-Glass time series. The proposed model was also compared with the quality of some "conventional" fuzzy models encountered in the literature.

• International Journal of Control, Automation, and Systems
• /
• v.3 no.2
• /
• pp.183-194
• /
• 2005

In this study, we introduce an advanced architecture of genetically optimized Hybrid Fuzzy Neural Networks (gHFNN) and develop a comprehensive design methodology supporting their construction. A series of numeric experiments is included to illustrate the performance of the networks. The construction of gHFNN exploits fundamental technologies of Computational Intelligence (CI), namely fuzzy sets, neural networks, and genetic algorithms (GAs). The architecture of the gHFNNs results from a synergistic usage of the genetic optimization-driven hybrid system generated by combining Fuzzy Neural Networks (FNN) with Polynomial Neural Networks (PNN). In this tandem, a FNN supports the formation of the premise part of the rule-based structure of the gHFNN. The consequence part of the gHFNN is designed using PNNs. We distinguish between two types of the linear fuzzy inference rule-based FNN structures showing how this taxonomy depends upon the type of a fuzzy partition of input variables. As to the consequence part of the gHFNN, the development of the PNN dwells on two general optimization mechanisms: the structural optimization is realized via GAs whereas in case of the parametric optimization we proceed with a standard least square method-based learning. To evaluate the performance of the gHFNN, the models are experimented with a representative numerical example. A comparative analysis demonstrates that the proposed gHFNN come with higher accuracy as well as superb predictive capabilities when comparing with other neurofuzzy models.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.53 no.6
• /
• pp.415-424
• /
• 2004

In this paper, we introduce and investigate a new category of rule-based fuzzy inference system based on Information Granulation(IG). The proposed rule-based fuzzy modeling implements system structure and parameter identification in the efficient form of “If..., then...” statements, and exploits the theory of system optimization and fuzzy implication rules. The form of the fuzzy rules comes with three types of fuzzy inferences: a simplified one that involves conclusions that are fixed numeric values, a linear one where the conclusion part is viewed as a linear function of inputs, and a regression polynomial one as the extended type of the linear one. By the nature of the rule-based fuzzy systems, these fuzzy models are geared toward capturing relationships between information granules. The form of the information granules themselves becomes an important design features of the fuzzy model. Information granulation with the aid of HCM(Hard C-Means) clustering algorithm hell)s determine the initial parameters of rule-based fuzzy model such as the initial apexes of the membership functions and the initial values of polynomial function being used in the Premise and consequence Part of the fuzzy rules. And then the initial Parameters are tuned (adjusted) effectively with the aid of the improved complex method(ICM) and the standard least square method(LSM). In the sequel, the ICM and LSM lead to fine-tuning of the parameters of premise membership functions and consequent polynomial functions in the rules of fuzzy model. An aggregate objective function with a weighting factor is proposed in order to achieve a balance between performance of the fuzzy model. Numerical examples are included to evaluate the performance of the proposed model. They are also contrasted with the performance of the fuzzy models existing in the literature.

• Journal of KIISE:Databases
• /
• v.37 no.4
• /
• pp.197-202
• /
• 2010

Recent studies on change detection for RDF data are focused on not only the structural difference but also the semantic-aware difference by computing the closure of RDF models. However, since these techniques which take into account the semantics of RDF model require both RDF models to be memory resident, or they use a forward-chaining strategy which computes the entire closure in advance, it is not efficient to apply them directly to detect changes in large RDF data. In this paper, we propose a scalable change detection technique for RDF data, which uses a backward-chaining inference based on relational database. Proposed method uses a new approach for RDF reasoning that computes only the relevant part of the closure for change detection in a relational database. We show that our method clearly outperforms the previous works through experiment using the real RDF from the bioinformatics domain.