• Journal of the Korean Data and Information Science Society
• /
• v.21 no.5
• /
• pp.831-839
• /
• 2010

Least squares support vector machine (LS-SVM) is a kernel trick gaining a lot of popularities in the regression and classification problems. We use LS-SVM to propose a iterative algorithm for a nonlinear generalized autoregressive conditional heteroscedasticity model in the mean (GARCH-M) model to estimate the mean and the conditional volatility of stock market returns. The proposed method combines a weighted LS-SVM for the mean and unweighted LS-SVM for the conditional volatility. In this paper, we show that nonlinear GARCH-M models have a higher performance than the linear GARCH model and the linear GARCH-M model via real data estimations.

• Journal of the Korean Society for Precision Engineering
• /
• v.10 no.2
• /
• pp.126-137
• /
• 1993

This paper presents an efficient method for the identification of motion error sources in NC machine tools by making use of the circular interpolation test, which is often used in estimating the motion accuracy of NC machine tools. Mathematical formulae are described for motion errors due to various kinds of error sources. Two identification formulae are proposed: one is based on the frequency analysis and the other is formulated with the weithted residual method. Motion error signal is classified into two patterns, mean errors(mean of CW and CCW test signals from mean errors). The sources of the mean errors are identified by using the frequency analysis technique and the sources of the deviation errors by the weighted residual formulaltion. A menu driven, user oriented, computer program is written to realize the full steps of the proposed identificationprocedure. Then, the identification method is applied to two NC machine tools.

• Korean Journal of Occupational Health Nursing
• /
• v.6 no.1
• /
• pp.5-13
• /
• 1997

For the purpose of investigating the subjective symptom of industrial fatigue, a questionnaire survey was carried out on 282 labor workers and 189 managerial workers who were employed at the manufacture of electronic products in two small scale industries. Checklist of industrial fatigue was composed of physical symptoms(10 items), mental symptoms(10 items), and sensory neurotic symptoms(10 items). The results were as follows : 1. Complain rate of fatigue was the highest in "eye strain" of physical symptom, "feel anxious about things" of mental symptom, and "feel stiffness in the neck or the shoulders" of sensory neurotic symptom in labor workers and managerial workers. 2. Managerial workers demonstrated II dominant type (mental or night work type), while labor workers demonstrated I dominant type of fatigue (general type). 3. Mean weighted score of fatigue complaints in labor workers (23.16) was significantly higher than that in managerial workers (20.34). 4. Mean weighted scores of fatigue complaints in male, 5~9 years of work duration, married, 4~5 hours of sleeping time, graduation of high school and college, and large of workload were significantly higher in labor workers than in managerial workers. 5. In poor work condition with temperature, ventilation, illumination and noise, the average weighted score was significantly higher in labor workers than in managerial workers.

• Journal of Korea Multimedia Society
• /
• v.3 no.6
• /
• pp.592-603
• /
• 2000

This paper is concerning off-line signature verification using a density function which is obtained by convolving the signature image with twelve-directional $5\times{5}$ gradient masks and the weighted fuzzy mean classifier. The twelve-directional density function based on Nevatia-Babu template gradient is related to the overall shape of a signature image and thus, utilized as a feature set. The weighted fuzzy mean classifier with the reference feature vectors extracted from only genuine signature samples is evaluated for the verification of freehand forgeries. The experimental results show that the proposed system can classify a signature whether it is genuine or forged with more than 98% overall accuracy even without any knowledge of varied freehand forgeries.

• Ocean Science Journal
• /
• v.42 no.2
• /
• pp.69-79
• /
• 2007

A vertical transect with 4 km length was established for the macrofaunal survey on the Chokchon macrotidal flat in Kyeonggi Bay, Incheon, Korea, 1994. Tidal elevation (m) and sediment mean grain size $(\phi)$ were inversely predicted by the transfer functions from the faunal assemblages. Three methods: weighted average using optimum value (WA), tolerance weighted version of the weighted average (WAT) and maximum likelihood calibration (MLC) were employed. Estimates of tidal elevation and mean grain size obtained by using the three different methods showed positively corresponding trends with the observations. The estimates of MLC were found to have the minimum value of sum of squares due to errors (SSE). When applied to the previous data $(1990\sim1992)$, each of three inference models exhibited high predictive power. This result implied there are visible relationships between species composition and faunas' critical environmental factors. Although a potential significance of the two major abiotic factors was re-affirmed, a weak tendency of biological interaction was detected from faunal distribution patterns across the flat. In comparison to the spatial and temporal patterns of the estimates, it was suggested that sediment characteristics were the primary factors regulating the distribution of macrofaunal assemblages, rather than tidal elevation, and the species composition may be sensitively determined by minute changes in substratum properties on a tidal flat.

• Journal of KIISE:Software and Applications
• /
• v.35 no.4
• /
• pp.255-264
• /
• 2008

This paper proposes a novel method using K-means and Non-negative matrix factorization (NMF) for topic -based multi-document summarization. NMF decomposes weighted term by sentence matrix into two sparse non-negative matrices: semantic feature matrix and semantic variable matrix. Obtained semantic features are comprehensible intuitively. Weighted similarity between topic and semantic features can prevent meaningless sentences that are similar to a topic from being selected. K-means clustering removes noises from sentences so that biased semantics of documents are not reflected to summaries. Besides, coherence of document summaries can be enhanced by arranging selected sentences in the order of their ranks. The experimental results show that the proposed method achieves better performance than other methods.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.15 no.3
• /
• pp.34-44
• /
• 2016

This study was initiated to evaluate the performance of methodologies to estimate the space mean speed(SMS) using the time mean speed(TMS) which was collected from the vehicle detection system(VDS) in expressways. To this end, the methodologies presented in prior studies were firstly summarized. It is very hard to achieve exact SMSs and TMSs due to mechanical and communication errors in the field. Thus, a microscopic traffic simulation model was utilized to evaluated the performance. As a result, the harmonic mean and volume-distance weighted harmonic mean were close to the SMS in the case in which the TMSs of individual vehicles were used. However, when the 30-second-interval aggregated TMS were used, the volume-distance weighted harmonic mean was outstanding. In this study, a radar detector was installed in the Joongbu expressway to collect the SMS. The trajectory of individual vehicles collected from the detector were used to calculate the SMS, which was compared with the estimates using other methodologies selected in this study. As a result, the volume-distance weighted mean was turned out to be close to the SMS. However, as the congestion becomes severe. the deviation between the two speed becomes bigger.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.16 no.10
• /
• pp.7061-7070
• /
• 2015

Multi-Response Surface Optimization aims at finding the optimal setting of input variables considering multiple responses simultaneously. The Weighted Mean Squared Error (WMSE) minimization approach, which imposes a different weight on the two components of mean squared error, squared bias and variance, first obtains WMSE for each response and then minimizes all the WMSEs at once. Most of the methods proposed for the WMSE minimization approach to date are classified into the prior preference articulation approach, which requires that a decision maker (DM) provides his/her preference information a priori. However, it is quite difficult for the DM to provide such information in advance, because he/she cannot experience the relationships or conflicts among the responses. To overcome this limitation, this paper proposes a posterior preference articulation method to the WMSE minimization approach. The proposed method first generates all (or most) of the nondominated solutions without the DM's preference information. Then, the DM selects the best one from the set of nondominated solutions a posteriori. Its advantage is that it provides an opportunity for the DM to understand the tradeoffs in the entire set of nondominated solutions and effectively obtains the most preferred solution suitable for his/her preference structure.

• Journal of the Korean Vacuum Society
• /
• v.8 no.3B
• /
• pp.315-321
• /
• 1999

In this paper, the SEM image processing system based on PC is designed, and a new noise reduction filtering algorithm is proposed. The SEM image obtained in semiconductor processing line is sensitive to noise, the weighted-D filter can remove uniform and Gaussian noise effectively, but can not remove impulse noise properly, A new improved filtering algorithm is proposed to reduce mixed-noise. The performance of the proposed filter is quantitatively evaluated by use of the normalized mean square errors (NMSE). The experimental results show that the performance of the proposed filter is obtained between 0.96 and 2.5 times better than that of weighted-D filter in NMSE evaluation.

• Proceedings of the Korean Statistical Society Conference
• /
• 2003.10a
• /
• pp.161-166
• /
• 2003

In this paper, a new form of generalized linear models is proposed. The proposed models consist of a distribution function of the mean response and a weighted linear combination of distribution functions of covariates. This form addresses a structural problem of the link function in the generalized linear models. Markov chain Monte Carlo methods are used to estimate the parameters within a Bayesian framework.