• Environmental and Resource Economics Review
• /
• v.13 no.2
• /
• pp.175-194
• /
• 2004

Volumes of research have been implemented to estimate and predict the oil price. These models, however, fail in accurately predicting oil price as a model composed of only a few observable variables is limiting. Unobservable variables and news that have been overlooked in past research, yet have a high likelihood of affecting the oil price. Hence, this paper analyses the news impact on the price. The standard GARCH model fails in capturing some important features of the data. The estimated news impact curve for the GARCH model, which imposes symmetry on the conditional variances, suggests that the conditional variance is underestimated for negative shocks and overestimated for positive shocks. Hence, this paper introduces the asymmetric or leverage volatility models, in which good news and bad news have different impact on volatility. They include the EGARCH, AGARCH, and GJR models. The empirical results showed that negative shocks introduced more volatility than positive shocks. Overall, the AGARCH and GJR were the best at capturing this asymmetric effect. Furthermore, the GJR model successfully revealed the shape of the news impact curve and was a useful approach to modeling conditional heteroscedasticity.

• Journal of applied mathematics & informatics
• /
• v.13 no.1_2
• /
• pp.441-446
• /
• 2003

Let { $X_{n}$ , n $\geq$ 1} be a sequence of independent and identically distributed random variables with a common continuous distribution function F(x) and probability density function f(x). Let $Y_{n}$ = max{ $X_1$, $X_2$, …, $X_{n}$ } for n $\geq$ 1. We say $X_{j}$ is an upper record value of { $X_{n}$ , n$\geq$1} if $Y_{j}$ > $Y_{j-1}$, j > 1. The indices at which the upper record values occur are given by the record times {u(n)}, n$\geq$1, where u(n) = min{j｜j>u(n-1), $X_{j}$ > $X_{u}$ (n-1), n$\geq$2} and u(1) = 1. We call the random variable X $\in$ Beta (1, c) if the corresponding probability cumulative function F(x) of x is of the form F(x) = 1-(1-x)$^{c}$ , c>0, 0$\leq$x$\leq$1. In this paper, we will give a characterization of the beta distribution of the first kind by considering conditional expectations of record values.s.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.9
• /
• pp.3204-3220
• /
• 2021

We perform the task of video object segmentation by incorporating a conditional random field (CRF) and convolutional neural networks (CNNs). Most methods employ a CRF to refine a coarse output from fully convolutional networks. Others treat the inference process of the CRF as a recurrent neural network and then combine CNNs and the CRF into an end-to-end model for video object segmentation. In contrast to these methods, we propose a novel higher-order CRF model to solve the problem of video object segmentation. Specifically, we use CNNs to establish a higher-order dependence among pixels, and this dependence can provide critical global information for a segmentation model to enhance the global consistency of segmentation. In general, the optimization of the higher-order energy is extremely difficult. To make the problem tractable, we decompose the higher-order energy into two parts by utilizing auxiliary variables and then solve it by using an iterative process. We conduct quantitative and qualitative analyses on multiple datasets, and the proposed method achieves competitive results.

• The Korean Journal of Applied Statistics
• /
• v.10 no.1
• /
• pp.151-161
• /
• 1997

An easy Monte Carlo Gibbs sampling approach is suggested for Bayesian analysis of cumulative logit models for ordinal polytomous data. Because in the cumulative logit model the posterior conditional distributions of parameters are not given in convenient forms for random sample generation, appropriate latent variables are introduced into the model so that in the new model all the conditional distributions are given in very convenient forms for implementation of the Gibbs sampler.

• Proceedings of the Korean Nuclear Society Conference
• /
• 1997.05a
• /
• pp.292-297
• /
• 1997

An alternating conditional expectation (ACE) algorithm, a kind of non-parametric regression method, is proposed to generate empirical correlations automatically. The ACE algorithm yields an optimal relationship between a dependent variable and multiple independent variables without any preprocessing and initial assumption on the functional forms. This algorithm is applied to a collection of 12,879 CHF data points for forced convective boiling hi vertical tubes to develop a new critical heat flux (CHF) correlation. The meat root mean square, and maximum errors of our new correlation are -0.558%, 12.5%, and 122.6%, respectively. Our CHF correlation represents the entire set of CHF data with an overall accuracy equivalent to or better than that of three existing correlations.

• International Area Studies Review
• /
• v.20 no.3
• /
• pp.123-144
• /
• 2016

This paper investigates the linkages between returns both in foreign exchange and stock markets, and uncertainties in two markets using daily data for the period of 16 July 2004 to 30 June 2014 in Nepalese economy. Four hypotheses are tested about how uncertainty influences the stock index and exchange rates. From the empirical results, a bivariate EGARCH-M model is the best to explain the volatility in the two markets. There is a negative relationship from the exchange rates return to stock price return. Empirical results do provide strong empirical confirmation that negative effect of stock index uncertainty and positive effect of exchange rates uncertainty on average stock index. GARCH-in-mean variables in AR modeling are significant and shows that there is positive effect of exchange rates uncertainty and negative effect of stock index uncertainty on average exchange rates. Stock index shocks have longer lived effects on uncertainty in the stock market than exchange rates shock have on uncertainly in the foreign exchange market. The effect of the last period's shock, volatility is more sensitive to its own lagged values.

• Journal of the Korean Data and Information Science Society
• /
• v.22 no.1
• /
• pp.89-97
• /
• 2011

The aims of this study were to propose a method of estimation for mean residual life function (MRLF) from conditional survival function using the Buckley and James's (1979) pseudo random variables, and then to assess the performance of the proposed method through the simulation studies. The mean squared error (MSE) of proposed method were less than those of the Cox's proportional hazard model (PHM) and Beran's nonparametric method for non-PHM case. Futhermore in the case of PHM, the MSE's of proposed method were similar to those of Cox's PHM. Finally, to evaluate the appropriateness of practical use, we applied the proposed method to the gastric cancer data. The data set consist of the 1, 192 patients with gastric cancer underwent surgery at the Department of Surgery, K-University Hospital.

• Journal of Korean Society for Quality Management
• /
• v.47 no.2
• /
• pp.271-281
• /
• 2019

Purpose: The purpose of this study is to suggest a more efficient distance measure taking into account the relationship between categorical variables for categorical data cluster analysis. Methods: In this study, the association-based dissimilarity was employed to calculate the distance between two categorical data observations and the distance obtained from the association-based dissimilarity was applied to the PAM cluster algorithms to verify its effectiveness. The strength of association between two different categorical variables can be calculated using a mixture of dissimilarities between the conditional probability distributions of other categorical variables, given these two categorical values. In particular, this method is suitable for datasets whose categorical variables are highly correlated. Results: The simulation results using several real life data showed that the proposed distance which considered relationships among the categorical variables generally yielded better clustering performance than the Hamming distance. In addition, as the number of correlated variables was increasing, the difference in the performance of the two clustering methods based on different distance measures became statistically more significant. Conclusion: This study revealed that the adoption of the relationship between categorical variables using our proposed method positively affected the results of cluster analysis.

• Fashion & Textile Research Journal
• /
• v.18 no.5
• /
• pp.606-616
• /
• 2016

We intend an empirical assessment of examining the differences in the appearance management behaviors and demographic variables among groups classified by the clothing consumption values. The questionnaires are administered to 493 female and male adults above 20 years old in Seoul, Gyeonggi-do, Daegu and Kyungpook regions. For analysis of data from 478 respondents, descriptive statistics, cluster analysis, Cronbach's ${\alpha}$, ANOVA, Duncan test and ${\chi}^2$ test were applied. We show the following results. First, Factor analyses were employed for the clothing consumption values and appearance management behaviors. Six factors were for clothing consumption values: Individuality, appearance attractive, social, functional, conditional and fashion clothing consumption value. Four factors were for appearance management behaviors: weight training, skin care, hair care, make-up and clothing selection. According to clothing consumption values, four groups were classified: the passive, functional, social, and active group. We did cluster analysis to the appearance management behaviors of weight training, skin care, hair care, make-up and clothing selection. Second, the social and active groups were more interested in individuality, appearance attractive, social, functional, conditional and fashion clothing value. And they were also more involved in appearance management behaviors. Third, among the demographic variables, the single and female in 20s and 30s with higher level of education belonged to the active group. In this contribution, we find significant differences in the appearance management behavior and demographic variables classified by the clothing consumption values.

• Journal of Korean Society for Quality Management
• /
• v.26 no.2
• /
• pp.39-50
• /
• 1998

A screening procedure, where one or more correlated variables are used for screeing, is reviewed from the point of statistical hypothesis testing. Without assuming a specific probability model for the joint distribution of the performance and screening variables, some principles are provided to establish the best screeing region. A, pp.ication examples are provided for two cases; ⅰ) the case where the performance variable is dichotomous and ⅱ) the case where the performance variable is continuous. In case ⅰ), a normal model is assumed for the conditional distribution of the screening variable given the performance variable. In case ⅱ), the performance and screening variables are assumed to be jointly normally distributed.