• Journal of Intelligence and Information Systems
• /
• v.26 no.1
• /
• pp.23-45
• /
• 2020

Big data is creating in a wide variety of fields such as medical care, manufacturing, logistics, sales site, SNS, and the dataset characteristics are also diverse. In order to secure the competitiveness of companies, it is necessary to improve decision-making capacity using a classification algorithm. However, most of them do not have sufficient knowledge on what kind of classification algorithm is appropriate for a specific problem area. In other words, determining which classification algorithm is appropriate depending on the characteristics of the dataset was has been a task that required expertise and effort. This is because the relationship between the characteristics of datasets (called meta-features) and the performance of classification algorithms has not been fully understood. Moreover, there has been little research on meta-features reflecting the characteristics of multi-class. Therefore, the purpose of this study is to empirically analyze whether meta-features of multi-class datasets have a significant effect on the performance of classification algorithms. In this study, meta-features of multi-class datasets were identified into two factors, (the data structure and the data complexity,) and seven representative meta-features were selected. Among those, we included the Herfindahl-Hirschman Index (HHI), originally a market concentration measurement index, in the meta-features to replace IR(Imbalanced Ratio). Also, we developed a new index called Reverse ReLU Silhouette Score into the meta-feature set. Among the UCI Machine Learning Repository data, six representative datasets (Balance Scale, PageBlocks, Car Evaluation, User Knowledge-Modeling, Wine Quality(red), Contraceptive Method Choice) were selected. The class of each dataset was classified by using the classification algorithms (KNN, Logistic Regression, Nave Bayes, Random Forest, and SVM) selected in the study. For each dataset, we applied 10-fold cross validation method. 10% to 100% oversampling method is applied for each fold and meta-features of the dataset is measured. The meta-features selected are HHI, Number of Classes, Number of Features, Entropy, Reverse ReLU Silhouette Score, Nonlinearity of Linear Classifier, Hub Score. F1-score was selected as the dependent variable. As a result, the results of this study showed that the six meta-features including Reverse ReLU Silhouette Score and HHI proposed in this study have a significant effect on the classification performance. (1) The meta-features HHI proposed in this study was significant in the classification performance. (2) The number of variables has a significant effect on the classification performance, unlike the number of classes, but it has a positive effect. (3) The number of classes has a negative effect on the performance of classification. (4) Entropy has a significant effect on the performance of classification. (5) The Reverse ReLU Silhouette Score also significantly affects the classification performance at a significant level of 0.01. (6) The nonlinearity of linear classifiers has a significant negative effect on classification performance. In addition, the results of the analysis by the classification algorithms were also consistent. In the regression analysis by classification algorithm, Naïve Bayes algorithm does not have a significant effect on the number of variables unlike other classification algorithms. This study has two theoretical contributions: (1) two new meta-features (HHI, Reverse ReLU Silhouette score) was proved to be significant. (2) The effects of data characteristics on the performance of classification were investigated using meta-features. The practical contribution points (1) can be utilized in the development of classification algorithm recommendation system according to the characteristics of datasets. (2) Many data scientists are often testing by adjusting the parameters of the algorithm to find the optimal algorithm for the situation because the characteristics of the data are different. In this process, excessive waste of resources occurs due to hardware, cost, time, and manpower. This study is expected to be useful for machine learning, data mining researchers, practitioners, and machine learning-based system developers. The composition of this study consists of introduction, related research, research model, experiment, conclusion and discussion.

• The Journal of the Petrological Society of Korea
• /
• v.21 no.2
• /
• pp.235-247
• /
• 2012

Peridotite xenoliths hosted by alkali basalts from South Korea occur in Baengnyeong Island, Jeju Island, Boeun, Asan, Pyeongtaek and Ganseong areas. K-Ar whole-rock ages of the basaltic rocks range from 0.1 to 18.9 Ma. The peridotites are dominantly lherzolites and magnesian harzburgites, and the constituent minerals are Fo-rich olivine ($Fo_{88.4-92.0}$), En-rich orthopyroxene, Di-rich clinopyroxene, and Cr-rich spinel (Cr# = 7.8-53.6). Hydrous minerals, such as pargasite and phlogopite, or garnet have not been reported yet. The Korean peridotites are residues after variable degree of partial melting (up to 26%) and melt extraction from fertile MORB mantle. However, some samples (usually refractory harzburgites) exhibit metasomatic enrichment of the highly incompatible elements, such as LREE. Equilibration temperatures estimated using two-pyroxene geothermometry range from ca. 850 to $1050^{\circ}C$. Sr and Nd isotopic compositions in clinopyroxene separates from the Korean peridotites show trends between depleted MORB-like mantle (DMM) and bulk silicate earth (BSE), which can be explained by secondary metasomatic overprinting of a precursor time-integrated depleted mantle. The Korean peridotite clinopyroxenes define mixing trends between DMM and EM2 end members on Sr-Pb and Nd-Pb isotopic correlation diagrams, without any corresponding changes in the basement. This is contrary to what we observe in late Cenozoic intraplate volcanism in East Asia which shows two distinct mantle sources such as a DMM-EM1 array for NE China including Baengnyeong Island and a DMM-EM2 array for Southeast Asia including Jeju Island. This observation suggests the existence of large-scale two distinct mantle domains in the shallow asthenosphere beneath East Asia. The Re-Os model ages on Korean peridotites indicate that they have been isolated from convecting mantle between ca. 1.8 and 1.9 Ga.

• Journal of The Korean Association For Science Education
• /
• v.36 no.6
• /
• pp.959-968
• /
• 2016

In this study, the Delphi Method was conducted to extract variables as effects of integrated education. Forty-six experts engaged in both the integrated education and research fields participated in this study. The Delphi survey was conducted for three rounds. In the first round, an open questionnaire was given asking variables possibly considered as effects of integrated education. In the second round, variables induced from analysis of the first survey results were given and the degree of agreement for each variable was determined according to the Likert scale. In the third round of the survey, mean, standard deviation, and the first and third quartile calculated using the results of the second survey were given to experts to determine their degree of assent. In addition, categories for variables were suggested. The degree of agreement for appropriateness of categorization and relative importance were determined As a result, a total of 18 variables were chosen except for career awareness. They were categorized according to their definition and properties into five categories: 'creativity' (flexible thinking, associative thinking, intuitive thinking, creative thinking), 'problem solving' (meta-cognition, problem recognition and solving, critical thinking, decision making ability, ability of knowledge application, knowledge and information processing skills), 'integrative perception and sensitivity' (concern and interest in various disciplines, understanding and acceptance of difference, integrative thinking), 'interpersonal relations' (communication skills, cooperation), and 'disciplinary literacy' (humanistic imagination, basic knowledge and literacy of each discipline, academic motivation). The degree of agreement was high in variables included in 'creativity' and 'problem solving' categories and the frequency of choosing the importance was high in variables included in 'integrative perception and sensitivity'. The educational implication related to implementation and practice of integrated education were discussed on the basis of results.

• Asia-Pacific Journal of Business Venturing and Entrepreneurship
• /
• v.15 no.2
• /
• pp.153-169
• /
• 2020

A new way of shopping based on virtual assistant, so called voice shopping, is drawing attention. The voice shopping market is growing around the world, and Korea is on the verge of full-scale commercialization of this new shopping. For the development of voice shopping-related industries, it is necessary to research on specific issues related to this new shopping methods, such as the quality of services, efficient processes tailored to new ways, and ways to build customer relationships. As part of such an attempt, the study seeks to determine the factors that affect consumers' perception and attitudes toward voice shopping. The study conducted the analysis based on survey response data of 171 online shopping users. In addition to the typical factors of the technology acceptability model(TAM) such as perceived usefulness and ease of use, the impact of perceived playfulness was included for analyzing the intention on the acceptance of voice shopping. In particular, this study focuses on the impact of user attributes. For the spread of voice shopping, it is necessary to set up a valid target customer and understand users for establishing an effective customer relationship. Therefore, this study tries to analyze how the perceptions on the voice shopping(perceived usefulness, ease of use, and perceived playfulness) are affected by users' attributes, such as user innovativeness and user knowledge level. The result of analysis shows that user innovativeness have a positive relationship with all of perceived usefulness, ease of use, and perceived playfulness. The user knowledge base, however, was not significant to all these three variables. The user knowledge base is shown to have a positive effect on user innovativeness which is the source of positively significant factor for the variable of the perceptions on the voice shopping. Meanwhile, among the variables of extended technology acceptance model, perceived usefulness and perceived playfulness have positive effects on the acceptance of voice shopping, while ease of use has no significant impact on the voice shopping acceptance. Ease of use has a positive relationship with perceived usefulness and playfulness. This study is meaningful in providing implications on the development of voice shopping platforms and related services, and establishment of customer relationship.

• Journal of the Korean Society of Food Science and Nutrition
• /
• v.37 no.11
• /
• pp.1435-1442
• /
• 2008

The purpose of this study was to identify the influence of consumption value on healthy functional food choice. Also, this study explored the role of health motivation, health concern, and food involvement as a moderating variable in the relationship between consumption value and healthy functional food choice. A total of 281 responses were collected using on-site survey (response rate 96.0%) from college students in Daegu, Gyeoungbuk Province. The questionnaire contained questions on consumption value, health motivation, health concern, food involvement, and purchasing intention of healthy functional food. The respondents rated the items using a 5-point scale from 1 (strongly disagree) to 5 (strongly agree). According to the confirmatory factor analysis, item evaluating using factor loading resulted in the retention of 25 consumption value items loading on seven factors, four health motivation items loading one factor, six health concern items loading on one factor, and four food involvement items loading on one factor with an internal consistency. Results of stepwise regression found that social value-I, emotional, functional, epistemic, and conditional values among consumption value determined the purchasing intention of healthy functional food. Results of hierarchical regression showed that health concern had a positive effect on the relationship between social value-I and purchasing intention of healthy functional food.

• Journal of Distribution Science
• /
• v.10 no.5
• /
• pp.37-44
• /
• 2012

In this paper, we investigate several factors that affect the locational decision of discount stores by using previous studies on the marketing area and the location of commercial facilities. We selected 21 primary variables that are expected to influence the decision of store location and, by factor analysis, grouped them into five underlying factors. Among these, the demographic factor, which shows the potential purchasing power level, had the greatest impact on the locational decision for the store. However, we found individual stores positioned according to unique locational characteristics in addition to the demographic factor. It means that we have to additionally consider if the vicinity of the market is based on any physical properties. Many previous studies proposed four decision factors for store location: the economic factor, the demographic factor, the land utilization factor, and traffic factor. However, the fivefold factors-our distinctive contribution-are more concrete and persuasive according to Korean reality. We show that location preference is based on the following criteria: (1) the area is densely populated, (2) houses stand close together, (3) residents have a high income level, (4) road traffic is developed and easy to access, and (5) public transportation is well developed. The demographic factor has the greatest impact on the location of a discount store. The number of households has a greater relevance to the demographic factor than does the individual consumer. Second, discount stores relatively prefer places where houses are located close together because such places offer easy access to the market. Third, a place whose residents have a high income level will be preferred, with its large cars and excellent traffic conditions. Fourth, a location would be highly rated if the roads around commercial facilities are well developed and their accessibility is good. Finally, discount stores must be located close to bus stops because female consumers, including housewives-the most important customers-evaluate stores based on distance. In this research, the variable of consumer attitude and preference was excluded, and the location factors of discount stores were analyzed according to a microscopic view through physical spatial data. In the future, the opening of new discount stores based on the five factors indicated above will require a comparatively shorter time from the first project feasibility analysis. In addition, the result of our study can be applied to the field of public policy for constructing and attracting large-scale distribution facilities.

• Sleep Medicine and Psychophysiology
• /
• v.8 no.1
• /
• pp.52-58
• /
• 2001

Objectives: An Increased level of psychophysiologic arousal and diminished physiologic flexibility would be observed in patients with panic disorder compared with a normal control group. We investigated the differences of psychophysiologic response between patients with panic disorder and normal control to examine this hypothesis. Methods: Ten Korean patients with panic disorder who met the diagnostic criteria of DSM-IV were compared with 10 normal healthy subjects. In psychological assessment, levels of anxiety and depression were evaluated by State-Trait Anxiety Inventory, Beck's Depression Inventory and Hamilton Rating Scale For Anxiety and Depression. Heart rate, respiration rate, electrodermal response, and electromyographic activity were measured by biofeedback system (J & J I-330 model) to determine psychophysiologic responses on autonomic nervous system. Stressful tasks included mental arithmetic, video game, hyperventilation, and talking about a stressful event. Psychophysiologic responses were measured according to the following procedures : baseline(3 min)-mental arithmetic (3 min)-rest (3 min)-video game (3 min)-rest (3 min)-hyperventilation (3 min)-rest (3 min)-talking about a stressful event (3 min). Results: The baseline level of anxiety and depression, electrodermal response (p=.017), electromyographic activity (p=.047) and heart rate (p=.049) of patients with panic disorder were significantly higher than those of the normal subject group. In electrodermal response, patient group had significantly higher startle response than the control group during hyperventilation (p=.001). Startle and recovery responses of heart rate in the patient group were significantly lower than responses in the control group during mental arithmetic (p=.007, p=.002). In electrodermal response of the patient group, startle response was significantly higher than recovery response during mental arithmetic (p=.000) and video game task (p=.021). Recovery response was significantly higher than startle response in respiratory response during hyperventilation. Conclusion: The results showed that patients with panic disorder had higher autonomic arousal than the control group, but the physiologic flexibility was variable. We suggest that it is helpful for treatment of panic disorder to decrease the level of autonomic arousal and to recover the physiologic flexibility in certain stressful event.

• KDI Journal of Economic Policy
• /
• v.12 no.2
• /
• pp.51-79
• /
• 1990

Research on technical efficiency, an important dimension of market performance, had received little attention until recently by most industrial organization empiricists, the reason being that traditional microeconomic theory simply assumed away any form of inefficiency in production. Recently, however, an increasing number of research efforts have been conducted to answer questions such as: To what extent do technical ineffciencies exist in the production activities of firms and plants? What are the factors accounting for the level of inefficiency found and those explaining the interindustry difference in technical inefficiency? Are there any significant international differences in the levels of technical efficiency and, if so, how can we reconcile these results with the observed pattern of international trade, etc? As the first in a series of studies on the technical efficiency of Korea's manufacturing industries, this paper attempts to answer some of these questions. Since the estimation of technical efficiency requires the use of plant-level data for each of the five-digit KSIC industries available from the Census of Manufactures, one may consture the findings of this paper as empirical evidence of technical efficiency in Korea's manufacturing industries at the most disaggregated level. We start by clarifying the relationship among the various concepts of efficiency-allocative effciency, factor-price efficiency, technical efficiency, Leibenstein's X-efficiency, and scale efficiency. It then becomes clear that unless certain ceteris paribus assumptions are satisfied, our estimates of technical inefficiency are in fact related to factor price inefficiency as well. The empirical model employed is, what is called, a stochastic frontier production function which divides the stochastic term into two different components-one with a symmetric distribution for pure white noise and the other for technical inefficiency with an asymmetric distribution. A translog production function is assumed for the functional relationship between inputs and output, and was estimated by the corrected ordinary least squares method. The second and third sample moments of the regression residuals are then used to yield estimates of four different types of measures for technical (in) efficiency. The entire range of manufacturing industries can be divided into two groups, depending on whether or not the distribution of estimated regression residuals allows a successful estimation of technical efficiency. The regression equation employing value added as the dependent variable gives a greater number of "successful" industries than the one using gross output. The correlation among estimates of the different measures of efficiency appears to be high, while the estimates of efficiency based on different regression equations seem almost uncorrelated. Thus, in the subsequent analysis of the determinants of interindustry variations in technical efficiency, the choice of the regression equation in the previous stage will affect the outcome significantly.

• Journal of the Korean Institute of Navigation
• /
• v.4 no.2
• /
• pp.45-79
• /
• 1980

The aim of this paper is to analyze the mechanics of price formation in the tramp shipping. For the purpose of this study, the main characteristics of tramp freight rates and the market is examined, and a brief examination of the nature ofthe costs of operation is given which are essential for the understanding of the functioning of shipping firms as well as for the understanding of developments in the tramp freight market. The demand and supply relationships in the market is also analysed in detail. Tramp shipping is an industry that has a market which functions under conditions that are not dissimilar to the theoretical model of perfect competition. However, it does notmean that tramp shipping market is a perfectly competitive market. It is apparent that this realworld competitive system has its imperfections, which means that the market for tramp shipping is near to being a perfectly competitive market on an internaitonal scale and it is freight are therefore subjext to the laws of supply and demand. In theory, the minimum freight rate in the short term is that at which the lowest cost vessels will lay-up in preference to operating, and is equal to the variable costs minus lay-up costs; and this would imply that in all times except those of full employment for ships there is a tendency for newer low-cost, and, probably, faster vessels to be driving the older high-cost vessels in the breaker's yards. In this case, shipowners may be reluctant to lay-up their ships becasue of obligations to crews, or because they would lose credibility with shippers or financiers, or simply because of lost prestige. Mainly, however, the decision is made on strictly economic grounds. When, for example, the total operating costs minus the likely freight earnings are greater than the cost of taking the ship out of service, maintaining it, and recommissioning it, then a ship may be considered for laying-up; shipowners will, in other words, run the ships at freight earnings below operating costs by as much as the cost of laying them up. As described above, the freight rates fixed on the tramp shipping market are subject to the laws of supply and demand. In other words, the basic properties of supply and demand are of significance so far as price or rate fluctuations in the tramp freight market are concerned. In connection with the same of the demand for tramp shipping services, the following points should be brone in mind: (a) That the magnitude of demand for sea transport of dry cargoes in general and for tramp shipping services in particular is increasing in the long run. (b) That owning to external factors, the demand for tramp shipping services is capable of varying sharphy at a given going of time. (c) The demad for the industry's services tends to be price inelastic in the short run. On the other hand the demand for the services offered by the individual shipping firm tends as a rule to be infinitely price elastic. In the meantime, the properties of the supply of the tramp shipping facilities are that it cannot expand or contract in the short run. Also, that in the long run there is a time-lag between entrepreneurs' decision to expand their fleets and the actual time of delivery of the new vessels. Thus, supply is inelastic and not capable of responding to demand and price changes at a given period of time. In conclusion, it can be safely stated that short-run changes in freight rates are a direct result of variations in the magnitude of demand for tramp shipping facilities, whilest the average level of freight rates is brought down to relatively low levels over prolonged periods of time.

• Journal of Korean Society of Transportation
• /
• v.21 no.4
• /
• pp.113-132
• /
• 2003

Using nonradial data envelopment analysis(DEA) under assumptions of strong disposability and variable returns scale, this paper annually estimates productive. technical and allocative efficiencies of three publicly-owned rail transit properties which are different in terms of organizational type: Seoul Subway Corporation(SSC, local public corporation), the Seoul Metropolitan Electrified Railways sector (SMESRS) of Korea National Railroad(the national railway operator controlled by the Ministry of Construction and Transportation(MOCT)), and Busan Urban Transit Authority (BUTA, the national authority controlled by MOCT). Using the estimation results of Tobit regression analysis. the paper next computes their true productive, true technical and true allocative efficiencies, which reflect only the impacts of internal factors such as production activity by removing the impacts of external factors such as an organizational type and a track utilization rate. And the paper also computes an organizational efficiency and annually gross efficiencies for each property. The paper then conceptualized that the property produces a single output(car-kilometers) using four inputs(labor, electricity, car & maintenance and track) and uses unbalanced panel data consisted of annual observations on SSC, SMESRS and BUTA. The results obtained from DEA show that, on an average, SSC is the most efficient property on the productive and allocative sides, while SMESRS is the most technically-efficient one. On the other hand. BUTA is the most efficient one on the truly-productive and allocative sides, while SMESRS on the truly-technical side. Another important result is that the differences in true efficiency estimates among the three properties are considerably smaller than those in efficiency estimates. Besides. the most cost-efficient organizational type appears to be a local public corporation represented by SSC, which is also the most grossly-efficient property. These results suggest that a measure to sort out the impacts of external factors on the efficiency of rail transit properties is required to assess fairly it, and that a measure to restructure (establish) an existing(a new) rail transit property into a local public corporation(or authority) is required to improve its cost efficiency.