• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.3 no.2
• /
• pp.119-126
• /
• 1983

The purposes of this study are to analyze/model the mode choice behavior of the regional traveler by express bus/express train and to offer useful source in deciding the public transportation policy. The data analyzed were trips of both modes from March, 1980 to November, 1981, between Seoul and other nineteen cities; the data were grouped as five groups according to the change of service variables. Service variables were travel time(unit: minute), cost(:won), average allocation time(:won), service hour(:hour), and dummy variables by mode. As model Logit Model with linear or log utility function were postulated. As the result of this study, some reseanable models were constructed at Model Type I(eq. 2. of this paper) based on the above data except the dummy. It was judged that the parameters calibrated by Group III and Group IV data in table 4, were optimal. Among the parameters, the parameter of travel cost was most reliable. There was a tendency preferring express bus to train in October and November. With the constructed model and Pivot-Point Method. the demand change of express train caused by the service variables' change could be forecasted over 99%.

• Journal of Korean Society of Transportation
• /
• v.21 no.2
• /
• pp.95-105
• /
• 2003

Because the basic assumptions of deterministic user equilibrium assignment that all network users have perfect information of network condition and determine their routes without errors are known to be unrealistic, several stochastic assignment models have been proposed to relax this assumption. However. it is not easy to solve such stochastic assignment models due to the probability distribution they assume. Also. in order to avoid all path enumeration they restrict the number of feasible path set, thereby they can not preciously explain the travel behavior when the travel cost is varied in a network loading step. Another problem of the stochastic assignment models is stemmed from that they use heuristic approach in attaining optimal moving size, due to the difficulty for evaluation of their objective function. This paper presents a logit-based stochastic assignment model and its solution algorithm to cope with the problems above. We also provide a stochastic user equilibrium condition of the model. The model is based on path where all feasible paths are enumerated in advance. This kind of method needs a more computing demand for running the model compared to the link-based one. However, there are same advantages. It could describe the travel behavior more exactly, and too much computing time does not require than we expect, because we calculate the path set only one time in initial step Two numerical examples are also given in order to assess the model and to compare it with other methods.

• Journal of Korean Society of Forest Science
• /
• v.110 no.4
• /
• pp.711-723
• /
• 2021

During the COVID-19 pandemic, social distancing affected daily life and leisure activities, including forest recreational activities. This study identified changes in people's participation in forest recreational activities and factors affecting their participation. It collected data from 1,000 samples through an online survey and analyzed it using a binary logit model with interaction terms. It was observed that there were decreases in the participation in visits to urban parks and green areas, outdoor activities in forests and mountains, and trips to mountain villages after the COVID-19 outbreak. People in their 40s, 50s, and 60s were more likely to decrease their participation in all kinds of forest recreational activities than those in their 20s. Moreover, higher household income earners were more likely to decrease their involvement in outdoor activities in forests and mountains. With respect to the place of residence, the residents in the Seoul metropolitan area were less likely to decrease their participation in trips to mountain villages than those outside this area. Thus, this study suggests that online forest recreation information services and forest management are needed to meet the demands of forest recreation for young generations and diversify the function of forests and rural areas as a safe leisurely space.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.4
• /
• pp.232-241
• /
• 2019

This study examines one of the conventional and controversial issues in modern finance. Specifically, this study identifies financial determinants of corporate R&D intensity for firms belonging to Korean Chaebols. Empirical estimation procedures are applied to derive more robust results of each hypothesis test. Static panel data, Tobit regression and stepwise regression models are employed to obtain significant financial factors of R&D expenditures, while logit, probit and complementary log-log regression models are used to detect financial differences between Chaebol firms and their counterparts not classified as Chaebols. Study results found the level of R&D intensity in the prior fiscal year, market-value based leverage ratio and firm size empirically showed their significance to account for corporate R&D intensity in the first hypothesis test, whereas the majority of explanatory variables had important power on a relative basis. Assuming that the current circumstances in the domestic capital market may necessitate gradual changes of Korean Chaebols in terms of their socio-economic function, the results of this study are expected to contribute to identifying financial antecedents that can be beneficial to attain optimal level of corporate R&D expenditures for Chaebol firms on a virtuous cycle.

• Journal of Intelligence and Information Systems
• /
• v.17 no.4
• /
• pp.241-254
• /
• 2011

Financial time-series forecasting is one of the most important issues because it is essential for the risk management of financial institutions. Therefore, researchers have tried to forecast financial time-series using various data mining techniques such as regression, artificial neural networks, decision trees, k-nearest neighbor etc. Recently, support vector machines (SVMs) are popularly applied to this research area because they have advantages that they don't require huge training data and have low possibility of overfitting. However, a user must determine several design factors by heuristics in order to use SVM. For example, the selection of appropriate kernel function and its parameters and proper feature subset selection are major design factors of SVM. Other than these factors, the proper selection of instance subset may also improve the forecasting performance of SVM by eliminating irrelevant and distorting training instances. Nonetheless, there have been few studies that have applied instance selection to SVM, especially in the domain of stock market prediction. Instance selection tries to choose proper instance subsets from original training data. It may be considered as a method of knowledge refinement and it maintains the instance-base. This study proposes the novel instance selection algorithm for SVMs. The proposed technique in this study uses genetic algorithm (GA) to optimize instance selection process with parameter optimization simultaneously. We call the model as ISVM (SVM with Instance selection) in this study. Experiments on stock market data are implemented using ISVM. In this study, the GA searches for optimal or near-optimal values of kernel parameters and relevant instances for SVMs. This study needs two sets of parameters in chromosomes in GA setting : The codes for kernel parameters and for instance selection. For the controlling parameters of the GA search, the population size is set at 50 organisms and the value of the crossover rate is set at 0.7 while the mutation rate is 0.1. As the stopping condition, 50 generations are permitted. The application data used in this study consists of technical indicators and the direction of change in the daily Korea stock price index (KOSPI). The total number of samples is 2218 trading days. We separate the whole data into three subsets as training, test, hold-out data set. The number of data in each subset is 1056, 581, 581 respectively. This study compares ISVM to several comparative models including logistic regression (logit), backpropagation neural networks (ANN), nearest neighbor (1-NN), conventional SVM (SVM) and SVM with the optimized parameters (PSVM). In especial, PSVM uses optimized kernel parameters by the genetic algorithm. The experimental results show that ISVM outperforms 1-NN by 15.32%, ANN by 6.89%, Logit and SVM by 5.34%, and PSVM by 4.82% for the holdout data. For ISVM, only 556 data from 1056 original training data are used to produce the result. In addition, the two-sample test for proportions is used to examine whether ISVM significantly outperforms other comparative models. The results indicate that ISVM outperforms ANN and 1-NN at the 1% statistical significance level. In addition, ISVM performs better than Logit, SVM and PSVM at the 5% statistical significance level.

• The Journal of Korean society of community based occupational therapy
• /
• v.2 no.1
• /
• pp.49-63
• /
• 2012

Objective : The purpose of study was to verify the clinical effect of a Korean Computer-based cognitive rehabilitation program(called CoTras) for recovering the visual perception function and ADL in stroke. Methods : A CBCRT was applied to 14 Stoke patients who rehabilitation professional medical treatment hospital. All participant were evaluated with four standardized assessment tolls(Motor-Free Visual Perception Test; MVPT, Korean version of Mini-Mental State Examination; MMSE-K, Assesment of Motor and Process Skills: AMPS) before and after the planned computer based cognitive rehabilitation sessions. Results : A significant effect was confirmed (p<.05) from the CBCRT which visual perception function. By each entry comparative result, visual memory, figure ground, visual close, spatial relation, visual discrimination, were the order of treatment. Neither was found any significant effect in improving process skills from AMPS. Conclusion : These results indicate that CoTras have effects on improving visual perception and ADL performance in stroke patients. Will be able to present with the fundamental data CoTras will be able to contribute to increase visual perception function & ADL performance to the stroke patient who has visual perception dysfunction.

• 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.25 no.3
• /
• pp.111-121
• /
• 2007

A transit assignment model can forecast the behaviors of transit users. thereby playing an important role In the evaluation of transit policies. Most existing transit assignment models are based on the models for passenger cars; therefore they cannot reflect the specific characteristics of transit modes. In addition most of the existing models are based on a single transit mode (bus or rail), and they cannot forecast the behaviors of transit users in a changing mass transportation system. The goal of this study is to overcome these problems with the exiting models and to develop a more realistic model. The newly developed model is based on mixed transit modes and is a stochastic model that can reflect the different preferences of each transit user for travel time and transfering. Data gathered from the Seoul metropolitan area's smart card are used to calibrate this model. This study is expected to be used for the evaluation of transportation policies and to attribute the development of transit revitalization strategies.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.31 no.5D
• /
• pp.671-677
• /
• 2011

A diversity of railway network function enhancement projects such as the double tracking, electrification, and direct operation have been actively executed to improve the railway service. When the new rapid transit is provided, how many people will use it instead of other transports? How will the railway choice behavior be changed? Accordingly, in this paper, the applicability of diverted travel demand forecast methods, by Revealed Preference(RP) and Stated Preference(SP) data was reviewed for Daegu metropolitan rail rapid transit service. As the result of combining RP and SP data, including the sequential and simultaneous approach, the total travel time and travel cost parameters are of the right sign and are highly significant. The simultaneous approach is more efficient in terms of the estimation of coefficients. In particular, methods to improve validity of the Mixed RP/SP models, when RP data is used proportionally, the diverted travel demand can be easily identified by railway fare and travel time service level. Therefore, it is considered that this will practically apply even in other regions as well as Daegu metropolitan railway.

• Journal of the Korean Society for Aviation and Aeronautics
• /
• v.20 no.1
• /
• pp.26-33
• /
• 2012

Competition between air transportation and railways has grown fiercer in major countries around the world with the rise of high-speed railways. In South Korea, air passenger travel has been rapidly decreasing since the initial launch of the Seoul-Pusan KTX line in 2004 and second opening that followed in 2010. Further expansion of the high-speed railway is expected. At present, research efforts to verify the validity of constructing an underwater express railway tunnel between Ho-nam and Jeju Island are taking place. Considering the possible high speed railway connection between Seoul and Jeju Island, this thesis has analyzed the choice behavior of existing passengers of the major and low-cost carriers. For this, Stated Preference (SP) research has been performed for three variables, including fare, travel time and the number of runs, to estimate the substitution rate of each of the three variables. Binomial Logit Model has been estimated with the obtained data. The estimation of the model has found that airline passengers of major and low-cost carriers are willing to pay approximately 7,200 KRW and 5,000 KRW, respectively, to reduce travel time by one hour. If the number of runs in one day increases, it has been estimated that the passengers are willing to pay additional fares of about 390 KRW and 30 KRW, respectively. On the other hand, the substitution rate between the number of runs and the travel time was found to be somewhat insignificant. If the construction of the Seoul-Jeju line progresses in the future, this study could be used as preliminary data for determining fares, travel time and the number of runs.