• The Korean Journal of Applied Statistics
• /
• v.30 no.1
• /
• pp.147-157
• /
• 2017

The test of transition probabilities in panel Markov chains are introduced. We deal with the hypotheses whether panel Markov chains have the same transition probabilities or not for all times. We suggest a LR test statistic for the test and its limit distribution is derived. We perform a simulation study to examine the limit distribution of test statistics when the number of the individuals are large.

• The Korean Journal of Applied Statistics
• /
• v.17 no.3
• /
• pp.547-556
• /
• 2004

This paper consider the testing problem of variance component for the unbalanced two-way error component model with nuisance parameter. We derive the one-sided LM test statistic for testing zero individual(time) effects assuming that the other time-specific(individual) effects are present. Using the Monte Carlo experiments, the computational more demanding LR test slightly underestimates the nominal size and has the low powers relative to LM test statistic.

• Communications for Statistical Applications and Methods
• /
• v.18 no.2
• /
• pp.219-227
• /
• 2011

In this paper, we proposed a generalized bivariate negative binomial distribution allowing heterogeneous dispersions on two dependent variables based on a trivariate reduction technique. In this model, we propose the score and LR tests for testing the equality of dispersions and compare the efficiencies of the proposed tests using a Monte Carlo study. The Monte Carlo study shows that the proposed score and LR tests prove to be an efficient test for the equality of dispersions in the view of the significance level and power. However, the score test is easier to compute than the LR test and it shows a slightly better performance than the LR test from the Monte Carlo study, we suggest the use of score tests for testing the equality of dispersions on two dependent variables. In addition, an empirical example is provided to illustrate the results.

• Management & Information Systems Review
• /
• v.28 no.4
• /
• pp.93-108
• /
• 2009

This study examines the asymmetric volatility in the Korean bond market and stock market by using the KTB Prime Index and KOSPI. Because accurate estimation and forecasting of volatility is essential before investing assets, it is important to understand the asymmetric response of volatility in bond market. Therefore I investigate the existence of asymmetric volatility in Korean bond market unlike the previous studies which mainly focused on stock returns. The main results of the empirical analysis with GARCH and GJR-GARCH model are as follow. At first, it exists the asymmetric volatility on KOSPI returns like the previous studies. Also, I find that the GJR-GARCH is more suitable one than GARCH model for forecasting volatility. Second, it does not exist the asymmetric volatility on KTB Prime Index returns. This result is showed by that using the GARCH model for forecasting volatility in bond market is sufficient.

• KIPS Transactions on Software and Data Engineering
• /
• v.9 no.5
• /
• pp.153-160
• /
• 2020

Energy Consumption Predictions for Industries has a prominent role to play in the energy management and control system as dynamic and seasonal changes are occurring in energy demand and supply. This paper introduces and explores the steel industry's predictive models of energy consumption. The data used includes lagging and leading reactive power lagging and leading current variable, emission of carbon dioxide (tCO2) and load type. Four statistical models are trained and tested in the test set: (a) Linear Regression (LR), (b) Radial Kernel Support Vector Machine (SVM RBF), (c) Gradient Boosting Machine (GBM), and (d) Random Forest (RF). Root Mean Squared Error (RMSE), Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE) are used for calculating regression model predictive performance. When using all the predictors, the best model RF can provide RMSE value 7.33 in the test set.

• Journal of Korea Water Resources Association
• /
• v.47 no.11
• /
• pp.1039-1050
• /
• 2014

The goal of this study is to demonstrate the diversity of model performance for various climatic elements and indicators. We evaluated the skills of the most advanced 17 General Circulation Models (GCMs) i.e., CMIP5 (Climate Model Inter-comparison project, phase 5) climate models in reproducing retrospective climatology from 1950 to 2000 over the Southeast US for the key climatic elements important in the hydrological and agricultural perspectives (i.e., precipitation, maximum and minimum temperature, and wind speed). The biases of raw CMIP5 GCMs were estimated for 16 different climatic indicators that imply mean climatology, temporal variability, extreme frequency, etc. using a grid-based observational dataset as reference. Based on the error (RMSE) and correlation (R) of GCM outputs, the error-based GCM ranks were assigned on average over the indicators. Overall, the GCMs showed much better accuracy in representing mean climatology of temperature comparing to other elements whereas few GCM showed acceptable skills for precipitation. It was also found that the model skills and ranks would be substantially different by the climatic elements, error statistics applied for evaluation, and indicators as well. This study presents significance of GCM uncertainty and the needs of considering rational strategies for climate model evaluation and selection.