• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제13권3호
• /
• pp.186-199
• /
• 2013

We pose pattern classification as a density estimation problem where we consider mixtures of generative models under partially labeled data setups. Unlike traditional approaches that estimate density everywhere in data space, we focus on the density along the decision boundary that can yield more discriminative models with superior classification performance. We extend our earlier work on the recursive estimation method for discriminative mixture models to semi-supervised learning setups where some of the data points lack class labels. Our model exploits the mixture structure in the functional gradient framework: it searches for the base mixture component model in a greedy fashion, maximizing the conditional class likelihoods for the labeled data and at the same time minimizing the uncertainty of class label prediction for unlabeled data points. The objective can be effectively imposed as individual mixture component learning on weighted data, hence our mixture learning typically becomes highly efficient for popular base generative models like Gaussians or hidden Markov models. Moreover, apart from the expectation-maximization algorithm, the proposed recursive estimation has several advantages including the lack of need for a pre-determined mixture order and robustness to the choice of initial parameters. We demonstrate the benefits of the proposed approach on a comprehensive set of evaluations consisting of diverse time-series classification problems in semi-supervised scenarios.

• 한국방사성폐기물학회:학술대회논문집
• /
• 한국방사성폐기물학회 2009년도 학술논문요약집
• /
• pp.186-186
• /
• 2009

The characterization strategy of fracture networks are classified into a deterministic or statistical characterization according to the type of required information. A deterministic characterization is most efficient for a sparsely fractured system, while the statistics are sufficient for densely fractured rock. In this study, the ensemble mean and variability of the effective connectivity is systematically analyzed with various density values for different network structures of a power law size distribution. The results of high resolution Monte Carlo analyses show that statistical characteristics can be a necessary information to determine the transport properties of a fracture system when fracture density is greater than a percolation threshold. When the percolation probability (II) approaches unity with increasing fracture density, the effective connectivity of the network can be safely estimated using statistics only (sufficient condition). It is inferred from conditional simulations that deterministic information for main pathways can reduce the uncertainty in estimation of system properties when the network becomes denser. Overall results imply that most pathways need to be identified when II < 0.5 statistics are sufficient when II $\rightarrow$ 1 and statistics are necessary and the identification of main pathways can significantly reduce the uncertainty in estimation of transport properties when 0.5$\ll$1. It is suggested that the proper estimation of the percolation probability of a fracture network is a prerequisite for an appropriate conceptualization and further characterization.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권12호
• /
• pp.6009-6027
• /
• 2019

Support vector machine (SVM) classifiers have been widely used for object detection. These methods usually locate the object by finding the region with maximal score in an image. With bag-of-features representation, the SVM score of an image region can be written as the sum of its inside feature-weights. As a result, the searching process can be executed efficiently by using strategies such as branch-and-bound. However, the feature-weight derived by optimizing region classification cannot really reveal the category knowledge of a feature-point, which could cause bad localization. In this paper, we represent a region in an image by a collection of local feature-points and determine the object by the region with the maximum posterior probability of belonging to the object class. Based on the Bayes' theorem and Naive-Bayes assumptions, the posterior probability is reformulated as the sum of feature-scores. The feature-score is manifested in the form of the logarithm of a probability ratio. Instead of estimating the numerator and denominator probabilities separately, we readily employ the density ratio estimation techniques directly, and overcome the above limitation. Experiments on a car dataset and PASCAL VOC 2007 dataset validated the effectiveness of our method compared to the baselines. In addition, the performance can be further improved by taking advantage of the recently developed deep convolutional neural network features.

• Structural Engineering and Mechanics
• /
• 제75권6호
• /
• pp.771-784
• /
• 2020

Reliability analysis techniques combining with various surrogate models have attracted increasing attention because of their accuracy and great efficiency. However, they primarily focus on the structures with continuous response, while very rare researches on the reliability analysis for structures with discontinuous response are carried out. Furthermore, existing adaptive reliability analysis methods based on importance sampling (IS) still have some intractable defects when dealing with small failure probability, and there is no related research on reliability analysis for structures involving discontinuous response and small failure probability. Therefore, this paper proposes a novel reliability analysis method called AGPC-IS for such structures, which combines adaptive Gaussian process classification (GPC) and adaptive-kernel-density-estimation-based IS. In AGPC-IS, an efficient adaptive strategy for design of experiments (DoE), taking into consideration the classification uncertainty, the sampling uniformity and the regional classification accuracy improvement, is developed with the purpose of improving the accuracy of Gaussian process classifier. The adaptive kernel density estimation is introduced for constructing the quasi-optimal density function of IS. In addition, a novel and more precise stopping criterion is also developed from the perspective of the stability of failure probability estimation. The efficiency, superiority and practicability of AGPC-IS are verified by three examples.

• 대한산업공학회지
• /
• 제38권3호
• /
• pp.191-197
• /
• 2012

One-class classification (OCC) is one of the recent growing areas in data mining and pattern recognition. In the present study we examine a k-nearest neighbors data description (kNNDD) algorithm, one of the OCC algorithms widely used. In particular, we propose to use nonparametric estimation methods to determine the threshold of the kNNDD algorithm. A simulation study has been conducted to explore the characteristics of the proposed approach and compare it with the existing approach that determines the threshold. The results demonstrate the usefulness and flexibility of the proposed approach.

• 전기학회논문지P
• /
• 제58권3호
• /
• pp.241-248
• /
• 2009

This paper presents stochastic methodology based fault detection algorithm for induction motor systems. We measure current of healthy induction motors by means of hall sensor systems and then establish its probability distribution. We propose online probability density estimation which is effective in real-time implementation due to its simplicity and low computational burden. In addition, we accomplish theoretical analysis of the proposed estimation to demonstrate its convergence property by using statistical convergence and system stability theories. We apply our fault detection approach to three-phase induction motors and achieve real-time experiment for evaluating its reliability and practicability in industrial fields.

• Communications for Statistical Applications and Methods
• /
• 제11권2호
• /
• pp.381-397
• /
• 2004

The marginal likelihood has become an important tool for model selection in Bayesian analysis because it can be used to rank the models. We discuss the marginal likelihood for Poisson regression models that are potentially useful in small area estimation. Computation in these models is intensive and it requires an implementation of Markov chain Monte Carlo (MCMC) methods. Using importance sampling and multivariate density estimation, we demonstrate a computation of the marginal likelihood through an output analysis from an MCMC sampler.

• Journal of the Korean Data and Information Science Society
• /
• 제14권3호
• /
• pp.687-696
• /
• 2003

This article presents a new family of the estimating functions related with minimum distance estimations, and discusses its relationship to the family of the minimum density power divergence estimating equations. Two representative minimum distance estimations; the minimum $L_2$ distance estimation and the minimum Hellinger distance estimation are studied in the light of the theory of estimating equations. Despite of the desirable properties of minimum distance estimations, they are not widely used by general researchers, because theories related with them are complex and are hard to be computationally implemented in real problems. Hopefully, this article would be a help for understanding the minimum distance estimations better.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2001년도 ICCAS
• /
• pp.33.5-33
• /
• 2001

We consider here the problem of spectral estimation of multidimensional wide sense stationary (WSS) random process. A method, employing a special difference equation of correlation function, is proposed to solve the problem of multidimensional spectral estimation. In this approach, the special difference equation of correlation function is derived by modal decomposition method. Maximum likelihood estimator and Kalman filter are used to estimate the model parameters of the difference equation and the decomposed spectral residues. An algorithm is presented to estimate the multidimensional spectral density. According to the result of the simulation, these methods are feasible to estimate the spectral density of WSS process, which is realized by finite dimensional multivariable lineal system driven by white noise.

• Communications for Statistical Applications and Methods
• /
• 제24권1호
• /
• pp.1-13
• /
• 2017

This manuscript summarized advances in bootstrap methods for long-range dependent time series data. The stationary linear long-memory process is briefly described, which is a target process for bootstrap methodologies on time-domain and frequency-domain in this review. We illustrate time-domain bootstrap under long-range dependence, moving or non-overlapping block bootstraps, and the autoregressive-sieve bootstrap. In particular, block bootstrap methodologies need an adjustment factor for the distribution estimation of the sample mean in contrast to applications to weak dependent time processes. However, the autoregressive-sieve bootstrap does not need any other modification for application to long-memory. The frequency domain bootstrap for Whittle estimation is provided using parametric spectral density estimates because there is no current nonparametric spectral density estimation method using a kernel function for the linear long-range dependent time process.