• The Korean Journal of Applied Statistics
• /
• v.20 no.1
• /
• pp.53-60
• /
• 2007

Conditionally heteroscedastic time series models such as GARCH processes have frequently provided useful approximations to the real aspects of financial time series. It is not uncommon that financial time series exhibits near non-stationary, say, integrated phenomenon. For stationary GARCH processes, a shock to the current conditional variance will be exponentially converging to zero and thus asymptotically negligible for the future conditional variance. However, for the case of integrated process, the effect will remain for a long time, i.e., we have a persistent effect of a current shock on the future observations. We are here concerned with providing empirical evidences of persistent GARCH(1,1) for various fifteen domestic financial time series including KOSPI, KOSDAQ and won-dollar exchange rate. To this end, kurtosis and Integrated-GARCH(1,1) fits are reported for each data.

• The Korean Journal of Applied Statistics
• /
• v.33 no.4
• /
• pp.467-482
• /
• 2020

We investigate a monitoring procedure for the early detection of parameter changes in location-scale time series models. We introduce a detector for monitoring procedure based on modified residual cumulative sum (CUSUM). The asymptotic properties of the monitoring procedure are established under the null and alternative hypotheses. Simulation results and data analysis are also provided for illustration.

• MALSORI
• /
• no.44
• /
• pp.105-113
• /
• 2002

In Korean, each digit is monosyllable and some pairs are known to have high confusability, causing performance degradation of connected digit recognition systems. To improve the performance, in this paper, we employ various discriminant analyses (DA) including Linear DA (LDA), Weighted Pairwise Scatter LDA WPS-LDA), Heteroscedastic Discriminant Analysis (HDA), and Maximum Likelihood Linear Transformation (MLLT). We also examine several combinations of various DA for additional performance improvement. Experimental results show that applying any DA mentioned above improves the string accuracy, but the amount of improvement of each DA method varies according to the model complexity or number of mixtures per state. Especially, more than 20% of string error reduction is achieved by applying MLLT after WPS-LDA, compared with the baseline system, when class level of DA is defined as a tied state and 1 mixture per state is used.

• Journal of Korean Society for Quality Management
• /
• v.31 no.2
• /
• pp.165-182
• /
• 2003

The fatigue has been considered to the most important failure mode where optimal design or reliability prediction of the machinery in aircraft, atomic reactors, and structure systems, etc., is required. When the statistical analysis of fatigue life data is performed, some difficulties are present because of the following facts : nonlinear relationship, heteroscedastic data, large scatter in the data, censored data (runouts), and existence of fatigue limit. To find the S-N curve models that characterize fatigue strength better, this research compares existing fatigue curve models developed recently in terms of the residual mean square and the estimate of fatigue limit, etc. for various fatigue data sets.

• The Korean Journal of Applied Statistics
• /
• v.15 no.2
• /
• pp.311-321
• /
• 2002

Transfer function model(TFM) capturings conditional heteroscedastic pattern is introduced to analyze stochastic regression relationship between the two time series. Nonlinear ARCH concept is incorporated into the TFM via threshold ARCH and beta- ARCH models. Steps for statistical analysis of the proposed model are explained along the lines of the Box & Jenkins(1976, ch. 10). For illustration, dynamic analysis between KOSPI and NASDAQ is conducted from which it is seen that threshold ARCH performs the best.

• Journal of the Korean Statistical Society
• /
• v.16 no.2
• /
• pp.92-101
• /
• 1987

Our interest is to access in some way teh relative odds or probability that a multivariate observation Z belongs to one of k multivariate normal populations with unequal covariance matrices. We derived the empirical Bayes posterior odds ratio for the classification rule when population parameters are unknown. It is a generalization of the posterior odds ratio suggested by Gelsser (1964). The classification rule does not have complicated distribution theory which a large variety of techniques from the sampling viewpoint have. The proposed posterior odds ratio is compared to the Gelsser's posterior odds ratio through a Monte Carlo study. The results show that the empiricla Bayes posterior odds ratio, in general, performs better than the Gelsser's. Especially, for large dimension of Z and small training sample, the performance is prominent.

• Journal of environmental and Sanitary engineering
• /
• v.23 no.4
• /
• pp.83-93
• /
• 2008

The non-linear least squares model(NLSM) has long been the standard technique used by hydrologists for constructing rating curves. The reasons for its adaptation are vague, and its appropriateness as a method of describing discharge measurement uncertainty has not been well investigated. It is shown in this paper that the classical method of NLSM can model only a very limited class of variance heterogeneity. Furthermore, this lack of flexibility often leads to unaccounted heteroscedasticity, resulting in dubious values for the rating curve parameters and estimated discharge. By introducing a heteroscedastic maximum likelihood model(HMLM), the variance heterogeneity is treated more generally. The maximum likelihood model stabilises the variance better than the NLSM approach, and thus is a more robust and appropriate way to fit a rating curve to a set of discharge measurements.

• Communications for Statistical Applications and Methods
• /
• v.18 no.6
• /
• pp.809-816
• /
• 2011

A tree-indexed autoregressive(AR) process is a time series defined on a tree which is generated by a branching process and/or a deterministic splitting mechanism. This short article is concerned with conditional heteroscedastic structure of the tree-indexed AR models. It has been usual in the literature to analyze conditional mean structure (rather than conditional variance) of tree-indexed AR models. This article pursues to identify quadratic conditional heteroscedasticity inherent in various tree-indexed AR models in a unified way, and thus providing some perspectives to the future works in this area. The identical conditional variance of sisters sharing the same mother will be referred to as the branching heteroscedasticity(BH, for short). A quasilikelihood but preliminary estimation of the quadratic BH is discussed and relevant limit distributions are derived.

• Journal of the Korean Data and Information Science Society
• /
• v.16 no.2
• /
• pp.271-281
• /
• 2005

Zokian(1993) and Li and Li(1996) developed TARCH(Threshold ARCH) model, considering the asymmetries in volatility. The models are based on Engle(1982)'s ARCH model and Bollerslev(1986)'s GARCH model. However, two TARCH models can be expressed a common model through Box Cox Power transformation, which was used by Higgins and Bera(1992) for developing NARCH(nonlinear ARCH) model. This article shows the PTARCH(Power transformation TARCH) model is necessary in some condition, and it checks the fact that PTARCH model has better performance comparing estimates and RMSE(Root Mean Square Error) with those of Zakoian's TARCH model and Li and Li's TARCH model. PTARCH model would give contribution in asymmetric study as well as heteroscedastic study.

• Communications for Statistical Applications and Methods
• /
• v.17 no.2
• /
• pp.255-261
• /
• 2010

A class of threshold-asymmetric GRACH(TGARCH, hereafter) models has been useful for explaining asymmetric volatilities in the field of financial time series. The cumulative impulse response function of a conditionally heteroscedastic time series often measures a degree of unstability in volatilities. In this article, a general form of the cumulative impulse response function of the TGARCH model is discussed. In particular, We present formula in their closed forms for the first two lower order models, viz., TGARCH(1, 1) and TGARCH(2, 2).