• The Korean Journal of Applied Statistics
• /
• v.33 no.1
• /
• pp.25-46
• /
• 2020

This paper presents ordinal probit semiparametric regression models using Bayesian Spectral Analysis Regression (BSAR) method. Ordinal probit regression is a way of modeling ordinal responses - usually more than two categories - by connecting the probability of falling into each category explained by a combination of available covariates using a probit (an inverse function of normal cumulative distribution function) link. The Bayesian probit model facilitates posterior sampling by bringing a latent variable following normal distribution, therefore, the responses are categorized by the cut-off points according to values of latent variables. In this paper, we extend the latent variable approach to a semiparametric model for the Bayesian ordinal probit regression with nonparametric functions using a spectral representation of Gaussian processes based BSAR method. The latent variable is decomposed into a parametric component and a nonparametric component with or without a shape constraint for modeling ordinal responses and predicting outcomes more flexibly. We illustrate the proposed methods with simulation studies in comparison with existing methods and real data analysis applied to a Korean National Health and Nutrition Examination Survey (KNHANES) 2016 for investigating nonparametric relationship between smoking behavior and coffee intake.

• The Korean Journal of Applied Statistics
• /
• v.33 no.1
• /
• pp.47-59
• /
• 2020

In this paper, we present the longitudinal Markov binary regression model with t-link function when its transition order is known or unknown. It is assumed that logit or probit models are considered in binary regression models. Here, t-link function can be used for more flexibility instead of the probit model since the t distribution approaches to normal distribution as the degree of freedom goes to infinity. A Markov regression model is considered because of the longitudinal data of each individual data set. We propose Bayesian method to determine the transition order of Markov regression model. In particular, we use the deviance information criterion (DIC) (Spiegelhalter et al., 2002) of possible models in order to determine the transition order of the Markov binary regression model if the transition order is known; however, we compute and compare their posterior probabilities if unknown. In order to overcome the complicated Bayesian computation, our proposed model is reconstructed by the ideas of Albert and Chib (1993), Kuo and Mallick (1998), and Erkanli et al. (2001). Our proposed method is applied to the simulated data and real data examined by Sommer et al. (1984). Markov chain Monte Carlo methods to determine the optimal model are used assuming that the transition order of the Markov regression model are known or unknown. Gelman and Rubin's method (1992) is also employed to check the convergence of the Metropolis Hastings algorithm.

• Journal of Korean Society of Transportation
• /
• v.31 no.6
• /
• pp.53-66
• /
• 2013

The major purpose of this study is to evaluate methodologies to predict the injury severity of pedestrian-vehicle collisions. Methodologies to be evaluated and compared in this study include Binary Logistic Regression(BLR), Ordered Probit Model(OPM), Support Vector Machine(SVM) and Decision Tree(DT) method. Valuable insights into applying methodologies to analyze the characteristics of pedestrian injury severity are derived. For the purpose of identifying causal factors affecting the injury severity, statistical approaches such as BLR and OPM are recommended. On the other hand, to achieve better prediction performance, heuristic approaches such as SVM and DT are recommended. It is expected that the outcome of this study would be useful in developing various countermeasures for enhancing pedestrian safety.

• The Korean Journal of Applied Statistics
• /
• v.31 no.3
• /
• pp.367-382
• /
• 2018

Prediction models for a corporate bond rating in existing studies have been developed using various models such as linear regression, ordered logit, and random forest. Financial characteristics help build prediction models that are expected to be contained in the assigning model of the bond rating agencies. However, the ranges of bond ratings in existing studies vary from 5 to 20 and the prediction models were developed with samples in which the target companies and the observation periods are different. Thus, a simple comparison of the prediction accuracies in each study cannot determine the best prediction model. In order to conduct a fair comparison, this study has collected corporate bond ratings and financial characteristics from 2013 to 2017 and applied prediction models to them. In addition, we applied the elastic-net penalty for the linear regression, the ordered logit, and the ordered probit. Our comparison shows that data-driven variable selection using the elastic-net improves prediction accuracy in each corresponding model, and that the random forest is the most appropriate model in terms of prediction accuracy, which obtains 69.6% accuracy of the exact rating prediction on average from the 5-fold cross validation.

• The Korean Journal of Applied Statistics
• /
• v.30 no.5
• /
• pp.775-791
• /
• 2017

We frequently encounter binary data in real life. Logistic, Probit, Cauchit, Complementary log-log models are often used for binary data analysis. In order to analyze binary data, Liu (2004) proposed a Robit model, in which the inverse of cdf of the Student's t distribution is used as a link function. Kim et al. (2008) also proposed a generalized t-link model to make the binary regression model more flexible. The more flexible skewed distributions allow more flexible link functions in generalized linear models. In the sense, we propose a binary data regression model using skewed generalized t distributions introduced in Theodossiou (1998). We implement R code of the proposed models using the glm function included in R base and R sgt package. We also analyze Pima Indian data using the proposed model in R.

• The Journal of the Korea Contents Association
• /
• v.14 no.12
• /
• pp.973-988
• /
• 2014

This study investigated one of the contemporary issues in the Korean capital market and two hypotheses of concern were tested on the financial determinants of profitability for the firms belonging to the Korean chaebols during the era of the post-global financial turmoil. The first hypothesis applying conditional quantile regression (CQR) estimation provided the evidence that leverage ratio, fixed asset utilization, and foreign ownership among the nine quantitative explanatory variables, had overall statistical significance relative to the book-valued profitability measure, while additional variables such as a firm's size, fixed and a proxy for the type of exchange market showed their strong impacts on the market-valued profitability indicator. Concerning the formulated 'extended' DuPont system, only two components of EBITDAEBIT and EMULTIPLIER revealed their prominent influence on ROE (Return on Equity) over the two tested periods (the years 2008 and 2012).

• International Journal of Highway Engineering
• /
• v.10 no.3
• /
• pp.47-56
• /
• 2008

Generally, accident exposure at intersections is relatively higher than that at roadway segments due to more possibility of merging, diverging, turning, crossing, and weaving maneuver. Furthermore, the traffic accident rate at intersections has been rapidly increasing since 1990's. Since there is more opportunity of conflict at unsignalized intersection, frequency and severity of traffic accident are more severe than signalized intersections. The purpose of the study is to analyze factors causing vehicle crashes and provide intersection design guidelines to improve intersection safety. For this study, vehicle to vehicle crash data of 116 rural 3 legs unsignalized were collected and field surveys were conducted for traffic and geometric conditions. Ordered probit models were developed to analyze the severity of crashes. It was found that weather, obstacles in minor roadsides, presence of major exclusive right lane, presence of major road crosswalk, difference between posted speed of major road and minor road, land-use around intersections, shoulder width of major road, ADT of major road are significant factors for intersection safety.

• Journal of Technology Innovation
• /
• v.18 no.1
• /
• pp.99-121
• /
• 2010

The goal of this paper is to analyze the effects of various firm characteristics on the probability for a firm to receive government’s financial supports for R&D. In the empirical analysis, a Probit model is estimated for the 2008 Korea Innovation Survey data. The main contribution of the paper is to investigate the distribution of R&D supports at the national level, instead of the program level. Especially, it is the first academic effort to evaluate the effects of regional and industrial variables. The results show that: (1) firm size and export increase the probability of receiving government’s R&D support; (2) variables measuring firms’ innovative ability, such as official designation as innovative firm, running R&D institute, number of R&D personnel, also have significantly positive effects; (3) firms in the chemical and automobile industries are more likely to receive R&D supports; and (4) firms in Teakyoung and Bukyoung regions are more likely to receive R&D supports.

• Journal of Digital Convergence
• /
• v.14 no.11
• /
• pp.173-179
• /
• 2016

This study was conducted to provide basic information for future Anseong local food policy and local food activation by finding the key factor determining consumer purchasing for Anseong local food. First, we conducted a survey and derived consumer purchasing attributes for the local food. Logistic regression analysis was performed to find the main factors that determine the consumers' purchase intention for Anseong local food out of such seven attributes as 'excellent quality', 'safety', 'good for health', 'activation of local economy', 'low price', 'accessibility', 'variety of items'. The results showed that the most influencing attributes on consumers' purchase decisions for Anseong local food were 'excellent quality' and 'low price' followed by 'accessibility' and 'activation of local economy'.

• Journal of Technology Innovation
• /
• v.20 no.3
• /
• pp.57-91
• /
• 2012

This paper investigates determinants of litigation success in the two distinctive types of patent litigations, ex-parte and inter-parte cases, which are brought in the process where a filed application becomes a valid patent right. We regress winning rates of patent applicants on the characteristics of firms, trials, patent lawyer, and patent itself, using a probit model with sample selections. The paper finds that the relative suit rate of a firm, time to be sued, changes of patent agents by applicants, and multiple agents among explanatory variables affect ex-parte reexamination and in-parte post-grant patent trials differently in the point of average marginal effects. These variables lower the probability of applicant's victory in the ex-parte cases, while they raise the probability in the inter-parte trials. However, the experience that agents represent applicants is a winning rate-increasing factor both in inter-parte and ex-parte reexamination, unexpectedly. This result cannot be applied to the entire domain of the variable, since sample selection effects are reflected in the result. The number of claim increases the winning probability of the applicant in the both types of patent litigations. This study has some limitations because it ignores the information on the legal person to which a patent agent belongs, and confined agent's experience to patent filing. We leave it future studies to investigate the effects of lawsuit experience of patent agent, and those of characteristics of the law firm to which individual patent lawyer is affiliated.