• Proceedings of the IEEK Conference
• /
• 2003.07e
• /
• pp.2307-2310
• /
• 2003

In this paper, a method for face recognition based on the wavelet packet decomposition is proposed. In the proposed method, the input image is decomposed by the 2-level wavelet packet transformation and then the face areas are defined by the Integral Projection technique applied to each of the 1-level subband images, HL and LH. After the defined face areas are divided into three areas, called top, bottom, and border, the mean and the variance of the three areas of the approximation image are computed, and the variance of the single predetermined face area for the rest of 15 detail images, from which the feature vectors of statistical measure are extracted. In this paper we use the wavelet packet decomposition, a generalization of the classical wavelet decomposition, to obtain its richer signal analysis features such as discontinuity in higher derivatives, self-similarity, etc. And we have shown that even with very simple statistical features such as mean values and variance we can make an excellent basis for face classification, if an appropriate probability distance is used.

• Journal of Korean Society for Quality Management
• /
• v.41 no.2
• /
• pp.233-248
• /
• 2013

Purpose: This study develops heuristic process capability indices (PCIs) using distribution-decomposition methods and evaluates the performances. The heuristic methods decompose the variation of a quality characteristic into upper and lower deviations and adjust the value of the PCIs using decomposed deviations in accordance with the skewness. The weighted variance(WV), new WV(NWV), scaled WV(SWV), and weighted standard deviation(WSD) methods are considered. Methods: The performances of the heuristic PCIs are investigated under the varied situations such as various skewed distributions, sample sizes, and specifications. Results: WV PCI is the best under the normal populations, WSD and SWV PCIs are the best under the low skewed populations, NWV PCI is the best under the moderate and high skewed populations. Conclusion: Comprehensive analysis shows that the NWV method is most adequate for a practical use.

• The Korean Journal of Applied Statistics
• /
• v.33 no.3
• /
• pp.281-296
• /
• 2020

Repeated outcomes from the same subjects are referred to as longitudinal data. Analysis of the data requires different methods unlike cross-sectional data analysis. It is important to model the covariance matrix because the correlation between the repeated outcomes must be considered when estimating the effects of covariates on the mean response. However, the modeling of the covariance matrix is tricky because there are many parameters to be estimated, and the estimated covariance matrix should be positive definite. In this paper, we consider analysis of multivariate longitudinal data via two modeling methodologies for the covariance matrix for multivariate longitudinal data. Both methods describe serial correlations of multivariate longitudinal outcomes using a modified Cholesky decomposition. However, the two methods consider different decompositions to explain the correlation between simultaneous responses. The first method uses enhanced linear covariance models so that the covariance matrix satisfies a positive definiteness condition; in addition, and principal component analysis and maximization-minimization algorithm (MM algorithm) were used to estimate model parameters. The second method considers variance-correlation decomposition and hypersphere decomposition to model covariance matrix. Simulations are used to compare the performance of the two methodologies.

• The Korean Journal of Financial Management
• /
• v.22 no.2
• /
• pp.189-212
• /
• 2005

This study examines the interactions between KTB spot and futures markets using the daily prices from March 4, 2002 to January 31, 2005. We use Granger causality test, impulse Response Analysis and Variance Decomposition through vector autoregressive analysis (VAR). However, considering the long-term relationships between the level variables of KTB spot and futures, we introduced Vector Error Correction Model. The main results are as follows. According to the results of Granger-causality test and impulse response analysis, we find that the yields of KTB forward have a great influence on the change of KTB spot but not vice versa. In terms of volatility analysis, there is no inter-dependence between KTB forward and spot markets. In the variance decomposition analysis we find that the short-term KTB forward has much more impact on the KTB spot market than the long-term KTB forward does. We think these results are meaningful for bond investors who are in charge of capital asset pricing valuation, risk management and international portfolio management.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.28 no.1
• /
• pp.11-23
• /
• 2003

This study Investigates the short and long-run dynamic relationships between the domestic and U.S. stock markets for the period of declining stock prices. It Is well known that the domestic stock market variations are largely caused by the U.S. stock market movements. Multivariate causal tty test Is utilized to examine the lead-lag relationships among four stock prices of KOSPI and KOSDAQ In the domestic part and DOWJONES and NASDAQ In the U.S. part. When the stock prices tend to decrease In the long run, It Is found that both KOSPI and KOSDAQ have closer relations with NASDAQ than DOWJONES. When both of domestic stock markets are severely fluctuate, bidirectional causal relationships appear to exist between NASDAQ and each of KOSPI and KOSDAQ. On the other hand. when the domestic stock markets are relatively stable, unidirectional causality Is found to exist between NASDAQ and each of KOSPI and KOSDAQ. which is explicitly validated by the analysis of variance decomposition.

• Journal of Korean Institute of Industrial Engineers
• /
• v.25 no.1
• /
• pp.111-124
• /
• 1999

This evapaper is toluate the forecasting performance of three neural network(NN) approaches against ARIMA model using the famous time series analysis competition data. The first NN approach is to analyze the second Makridakis (M2) Competition Data using Multilayer Perceptron (MLP) that has been the most popular NN model in time series analysis. Since it is recently known that MLP suffers from bias/variance dilemma, two approaches are suggested in this study. The second approach adopts Cascade Correlation Network (CCN) that was suggested by Fahlman & Lebiere as an alternative to MLP. In the third approach, a time series is separated into two series using Noise Filtering Network (NFN) that utilizes autoassociative memory function of neural network. The forecasts in the decomposition analysis are the sum of two prediction values obtained from modeling each decomposed series, respectively. Among the three NN approaches, Decomposition Analysis shows the best forecasting performance on the M2 Competition Data, and is expected to be a promising tool in analyzing socio-economic time series data because it reduces the effect of noise or outliers that is an impediment to modeling the time series generating process.

• Journal of Korea Trade
• /
• v.24 no.7
• /
• pp.93-106
• /
• 2020

Purpose - This paper aims to examine the environmental effects of South Korean foreign trade, and the changing relationship between industrial "three wastes" emissions and foreign trade. Design/methodology - Based on time series data of South Korean foreign trade and industrial "three wastes" from 2009 to 2019, a VAR model was used to analyze the long-term internal links and dynamic changes between foreign trade and environmental pollution. Findings - Variance decomposition analysis shows that for the three types of pollutants, self-impact contributes the most to the variance decomposition. It follows that South Korean foreign trade has a certain negative impact on the environment, and this impact has a certain sustainability. Originality/value - This paper contributes to the study on the relationship between foreign trade and environmental pollution. It theoretically proposes a coordinated development path for foreign trade development and green development based on the environmental impact of foreign trade, to provide a reference for the development of collaborative promotion.

• Nuclear Engineering and Technology
• /
• v.52 no.2
• /
• pp.287-295
• /
• 2020

Group method of data handling (GMDH) is considered one of the earliest deep learning methods. Deep learning gained additional interest in today's applications due to its capability to handle complex and high dimensional problems. In this study, multi-layer GMDH networks are used to perform uncertainty quantification (UQ) and sensitivity analysis (SA) of nuclear reactor simulations. GMDH is utilized as a surrogate/metamodel to replace high fidelity computer models with cheap-to-evaluate surrogate models, which facilitate UQ and SA tasks (e.g. variance decomposition, uncertainty propagation, etc.). GMDH performance is validated through two UQ applications in reactor simulations: (1) low dimensional input space (two-phase flow in a reactor channel), and (2) high dimensional space (8-group homogenized cross-sections). In both applications, GMDH networks show very good performance with small mean absolute and squared errors as well as high accuracy in capturing the target variance. GMDH is utilized afterward to perform UQ tasks such as variance decomposition through Sobol indices, and GMDH-based uncertainty propagation with large number of samples. GMDH performance is also compared to other surrogates including Gaussian processes and polynomial chaos expansions. The comparison shows that GMDH has competitive performance with the other methods for the low dimensional problem, and reliable performance for the high dimensional problem.

• KIEE International Transactions on Power Engineering
• /
• v.4A no.1
• /
• pp.33-41
• /
• 2004

Knowledge of partial discharge (PD) is important to accurately diagnose and predict the condition of insulation. The PD phenomenon is highly complex and seems to be random in its occurrence. This paper indicates the possible use of chaos theory for the recognition and distinction concerning PD signals. Chaos refers to a state where the predictive abilities of a systems future are lost and the system is rendered aperiodic. The analysis of PD using deterministic chaos comprises of the study of the basic system dynamics of the PD phenomenon. This involves the construction of the PD attractor in state space. The simulation results show that the variance of an orthogonal axis in an attractor of chaos theory increases according to the magnitude and the number of PDs. However, it is difficult to clearly identify the characteristics of the PDs. Thus, we calculated the magnitude on an orthogonal axis in an attractor using singular value decomposition (SVD) and principal component analysis (PCA) to extract the numerical characteristics. In this paper, we proposed the condition monitoring method for gas insulated switchgear (GIS) using SVD for efficient calculation of the variance. Thousands of simulations have proven the accuracy and effectiveness of the proposed algorithm.

• The Journal of Fisheries Business Administration
• /
• v.46 no.1
• /
• pp.93-107
• /
• 2015

This study aims to analyze causalities among Hairtail prices by distribution channel using a vector autoregressive model. This study applies unit-root test for stability of data, uses Granger causality test to know interaction among Hairtail Prices by distribution channel, and employes the vector autoregressive model to estimate statistical impacts among t-2 period variables used in model. Analyzing results of this study are as follows. First, ADF, PP, and KPSS tests show that the change rate of Hairtail price by distribution channel differentiated by logarithm is stable. Second, a Granger causality test presents that the producer price of Hairtail leads the wholesale price and then the wholesale price leads the consumer price. Third, the vector autoregressive model suggests that the change rate of Hairtail producer price of t-2 period variables statistically, significantly impacts change rates of own, wholesale, and consumer prices at current period. Fourth, the impulse response analysis indicates that impulse responses of the structural shocks with a respectively distribution channel of the Hairtail prices are relatively more powerful in own distribution channel than in other distribution channels. Fifth, a forecast error variance decomposition of the Hairtail prices points out that the own price has relatively more powerful influence than other prices.