• Proceedings of the Korea Contents Association Conference
• /
• 2009.05a
• /
• pp.179-184
• /
• 2009

Robust face recognition under various illumination environments is difficult to achieve. For face recognition robust to illumination changes, usually face images are normalized with respect to illumination as a preprocessing step before face recognition. The anisotropic smoothing-based illumination normalization method, known to be one of the best illumination normalization methods, cannot handle casting shadows. In this paper, we present an efficient illumination normalization method for face recognition. The proposed illumination normalization method separates the effect of illumination from eigenfaces and constructs an illumination-separated eigenface subspace. Then, an incoming face image is projected into the subspace and the obtained projected face image is rendered so that illumination effects including casting shadows are reduced as much as possible. Application to real face images shows the proposed illumination normalization method.

• Journal of Intelligence and Information Systems
• /
• v.22 no.1
• /
• pp.139-157
• /
• 2016

Bankruptcy involves considerable costs, so it can have significant effects on a country's economy. Thus, bankruptcy prediction is an important issue. Over the past several decades, many researchers have addressed topics associated with bankruptcy prediction. Early research on bankruptcy prediction employed conventional statistical methods such as univariate analysis, discriminant analysis, multiple regression, and logistic regression. Later on, many studies began utilizing artificial intelligence techniques such as inductive learning, neural networks, and case-based reasoning. Currently, ensemble models are being utilized to enhance the accuracy of bankruptcy prediction. Ensemble classification involves combining multiple classifiers to obtain more accurate predictions than those obtained using individual models. Ensemble learning techniques are known to be very useful for improving the generalization ability of the classifier. Base classifiers in the ensemble must be as accurate and diverse as possible in order to enhance the generalization ability of an ensemble model. Commonly used methods for constructing ensemble classifiers include bagging, boosting, and random subspace. The random subspace method selects a random feature subset for each classifier from the original feature space to diversify the base classifiers of an ensemble. Each ensemble member is trained by a randomly chosen feature subspace from the original feature set, and predictions from each ensemble member are combined by an aggregation method. The k-nearest neighbors (KNN) classifier is robust with respect to variations in the dataset but is very sensitive to changes in the feature space. For this reason, KNN is a good classifier for the random subspace method. The KNN random subspace ensemble model has been shown to be very effective for improving an individual KNN model. The k parameter of KNN base classifiers and selected feature subsets for base classifiers play an important role in determining the performance of the KNN ensemble model. However, few studies have focused on optimizing the k parameter and feature subsets of base classifiers in the ensemble. This study proposed a new ensemble method that improves upon the performance KNN ensemble model by optimizing both k parameters and feature subsets of base classifiers. A genetic algorithm was used to optimize the KNN ensemble model and improve the prediction accuracy of the ensemble model. The proposed model was applied to a bankruptcy prediction problem by using a real dataset from Korean companies. The research data included 1800 externally non-audited firms that filed for bankruptcy (900 cases) or non-bankruptcy (900 cases). Initially, the dataset consisted of 134 financial ratios. Prior to the experiments, 75 financial ratios were selected based on an independent sample t-test of each financial ratio as an input variable and bankruptcy or non-bankruptcy as an output variable. Of these, 24 financial ratios were selected by using a logistic regression backward feature selection method. The complete dataset was separated into two parts: training and validation. The training dataset was further divided into two portions: one for the training model and the other to avoid overfitting. The prediction accuracy against this dataset was used to determine the fitness value in order to avoid overfitting. The validation dataset was used to evaluate the effectiveness of the final model. A 10-fold cross-validation was implemented to compare the performances of the proposed model and other models. To evaluate the effectiveness of the proposed model, the classification accuracy of the proposed model was compared with that of other models. The Q-statistic values and average classification accuracies of base classifiers were investigated. The experimental results showed that the proposed model outperformed other models, such as the single model and random subspace ensemble model.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2005.06a
• /
• pp.822-827
• /
• 2005

Recently engineering systems problems become quite large and complicated. For those problems, design requirements are fairly complex. It is not easy to design such systems by considering only one discipline. Therefore, we need a design methodology that can consider various disciplines. Multidisciplinary Design Optimization (MDO) is an emerging optimization method to include multiple disciplines. So far, about seven MDO methodologies have been proposed for MDO. They are Multidisciplinary Feasible (MDF), Individual Feasible (IDF), All-at-Once (AAO), Concurrent Subspace Optimization (CSSO), Collaborative Optimization (CO), Bi-Level Integrated System Synthesis (BLISS) and Multidisciplinary Optimization Based on Independent Subspaces (MDOIS). In this research, the performances of the methods are evaluated and compared. Practical engineering problems may not be appropriate for fairness. Therefore, mathematical problems are developed for the comparison. Conditions for fair comparison are defined and the mathematical problems are defined based on the conditions. All the methods are coded and the performances of the methods are compared qualitatively as well as quantitatively.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.22 no.3
• /
• pp.515-523
• /
• 1997

A high resolution algorithm is presented for resolving multiple coherent spread spectrum signals that are incident on an equispaced linear array. Unlike the conventional noise-eigenvector based methods, this algorithm makes use of the signal eigenvectors of the array spectral density matrix that are associates with eigenvalues that are larger than the sensor noise level. Simulation results are shown to demonstate the high performance of the proposed approach in comparison with MUSIC in which coherent signal subspace method (CSM) is employed.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.22 no.2
• /
• pp.192-197
• /
• 2012

With the advent of DNA microarray technologies, researches for disease diagnosis has been actively in progress. In typical experiments using microarray data, problems such as the large number of genes and the relatively small number of samples, the inherent measurement noise and the heterogeneity across different samples are the cause of the performance decrease. To overcome these problems, a new method using functional modules (e.g. signaling pathways) used as markers was proposed. They use the method using an activity of pathway summarizing values of a member gene's expression values. It showed better classification performance than the existing methods based on individual genes. The activity calculation, however, used in the method has some drawbacks such as a correlation between individual genes and each phenotype is ignored and characteristics of individual genes are removed. In this paper, we propose a method based on the ensemble classifier. It makes weak classifiers based on feature vectors using subsets of genes in selected pathways, and then infers the final classification result by combining the results of each weak classifier. In this process, we improved the performance by minimize the search space through a filtering process using gene-gene interaction information and the mRMR filter. We applied the proposed method to a classifying the lung cancer, it showed competitive classification performance compared to existing methods.

• The Journal of the Acoustical Society of Korea
• /
• v.27 no.5
• /
• pp.221-228
• /
• 2008

To estimate array shape with reference sources in SONAR systems, nearfield signal modeling is required for the reference sources near a towed array. Array shape estimation method based on the nearfield signal modeling generally exploits the spatial covariance matrix of the received reference sources. Among those method, nearfield eigenvector method uses the eigenvector corresponding to the maximum eigenvalue as a steering vector of the reference source. In this paper, we propose a simplified subspace fitting method based on the nearfield signal modeling with spherical wave modeling. Furthermore, we analyze performance of the array shape estimation methods based on the nearfield signal modeling for various environments. The results of the numerical experiments indicate that the simplified subspace fitting method and the nearfield eigenvector method with single reference source shows almost similar performance. Furthermore, the simplified subspace fitting method with 2 reference sources consistently estimates the shape of the array regardless of the incident angle of the reference sources, whereas the nearfield eigenvector method cannot apply for the case of 2 reference sources.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.28 no.4C
• /
• pp.384-391
• /
• 2003

Blind channel identification and equalization attempt to identify the communication channel and to remove the inter-symbol interference caused by a communication channel without using any known trainning sequences. In this paper, we propose a blind adaptive channel identification and equalization algorithm with phase offset compensation for single-input multiple-output (SIMO) channel. It is based on the one-step forward multichannel linear prediction error method and can be implemented by an RLS algorithm. Phase offset problem, we use a blind adaptive algorithm called the constant modulus derotator (CMD) algorithm based on condtant modulus algorithm (CMA). Moreover, unlike many known subspace (SS) methods or cross relation (CR) methods, our proposed algorithms do not require channel order estimation. Therefore, our algorithms are robust to channel order mismatch.

• Journal of the Korean Data Analysis Society
• /
• v.20 no.6
• /
• pp.2733-2746
• /
• 2018

The two inverse regression estimation methods, SIR and SAVE to estimate the central space are computationally easy and are widely used. However, SIR and SAVE may have poor performance in finite samples and need strong assumptions (linearity and/or constant covariance conditions) on predictors. The two non-parametric estimation methods, MAVE and dMAVE have much better performance for finite samples than SIR and SAVE. MAVE and dMAVE need no strong requirements on predictors or on the response variable. MAVE is focused on estimating the central mean subspace, but dMAVE is to estimate the central space. This paper explores and compares four methods to explain the dimension reduction. Each algorithm of these four methods is reviewed. Empirical study for simulated data shows that MAVE and dMAVE has relatively better performance than SIR and SAVE, regardless of not only different models but also different distributional assumptions of predictors. However, real data example with the binary response demonstrates that SAVE is better than other methods.

• Communications for Statistical Applications and Methods
• /
• v.28 no.3
• /
• pp.267-279
• /
• 2021

A regression with multi-dimensional responses is quite common nowadays in the so-called big data era. In such regression, to relieve the curse of dimension due to high-dimension of responses, the dimension reduction of predictors is essential in analysis. Sufficient dimension reduction provides effective tools for the reduction, but there are few sufficient dimension reduction methodologies for multivariate regression. To fill this gap, we newly propose two fused slice-based inverse regression methods. The proposed approaches are robust to the numbers of clusters or slices and improve the estimation results over existing methods by fusing many kernel matrices. Numerical studies are presented and are compared with existing methods. Real data analysis confirms practical usefulness of the proposed methods.

• Journal of Korea Multimedia Society
• /
• v.18 no.11
• /
• pp.1261-1267
• /
• 2015

The problem of underwater target detection and classification has been attracted a substantial amount of attention and studied from many researchers for both military and non-military purposes. The difficulty is complicate due to various environmental conditions. In this paper, we study classifier ensemble methods for active sonar target classification to improve the classification performance. In general, classifier ensemble method is useful for classifiers whose variances relatively large such as decision trees and neural networks. Bagging, Random selection samples, Random subspace and Rotation forest are selected as classifier ensemble methods. Using the four ensemble methods based on 31 neural network classifiers, the classification tests were carried out and performances were compared.