• The Korean Journal of Applied Statistics
• /
• v.32 no.6
• /
• pp.795-835
• /
• 2019

Big data has been generated in various fields. Many companies have now tried to make profits by building a system capable of analyzing big data based on artificial intelligence (AI) techniques. Integrating AI technology has made analyzing and utilizing vast amounts of data increasingly valuable. In particular, demand forecasting with maximum accuracy is critical to government and business management in various fields such as finance, procurement, production and marketing. In this case, it is important to apply an appropriate model that considers the demand pattern for each field. It is possible to analyze complex patterns of real data that can also be enlarged by a traditional time series model or regression model. However, choosing the right model among the various models is difficult without prior knowledge. Many studies based on AI techniques such as machine learning and deep learning have been proven to overcome these problems. In addition, demand forecasting through the analysis of stereotyped data and unstructured data of images or texts has also shown high accuracy. This paper introduces important areas where demand forecasts are relatively active as well as introduces machine learning and deep learning techniques that consider the characteristics of each field.

• Journal of Korean Society of Transportation
• /
• v.28 no.1
• /
• pp.39-49
• /
• 2010

Although many studies have been carried out on the pattern of behavior of drivers which result from the provision of traffic information, there have been few detailed studies on the composition of message, method for message expression, timing of provision, point of provision, media for provision, changes by traffic condition, etc. This study was intended to provide an insight into the changes in the characteristics related to the provision of information by analyzing how the patterns of information utilization change depending on the traffic condition and reclassifying such patterns according to the characteristics of media. Unlike the existing studies, this study adopted the traffic condition, using rate of information media, and the correlation coefficient label as the basis for information media classification, and categorized them into passive utilization media, active utilization media, and past experience in order to ensure the statistical reasonability. The categorized using rate of information media and traffic condition was found to have a positive(+) correlation with the travel speed in the case of passive utilization media during both consecutive holidays(Korea's traditional Thanksgiving day) and weekends, but had a negative(-) correlation with the positive utilization media and past experience. The rate of decision to take a detour based on the past experience was high at the condition of congestion or slow during both consecutive holidays and weekends, but the rate of decision to take a detour through passive utilization media was high in a smooth traffic. In other words, if the traffic condition worsens, using rate of passive utilization media would be low while the diversion rate would be high which uses the active utilization media and past experience. Therefore, it should be established to suit the traffic condition and media characteristics for strategies of traffic distribution through drivers' diversion behavior on weekends and consecutive holidays.

• Journal of Korea Spatial Information System Society
• /
• v.12 no.1
• /
• pp.18-27
• /
• 2010

In u-GIS environment, GeoSensor environment requires that dynamic data captured from various sensors and static information in terms of features in 2D or 3D are fused together. GeoSensors, the core of this environment, are distributed over a wide area sporadically, and are collected in any size constantly. As a result, storage space could be exceeded because of restricted memory in DSMS. To solve this kind of problems, a lot of related studies are being researched actively. There are typically 3 different methods - Random Load Shedding, Semantic Load Shedding, and Sampling. Random Load Shedding chooses and deletes data in random. Semantic Load Shedding prioritizes data, then deletes it first which has lower priority. Sampling uses statistical operation, computes sampling rate, and sheds load. However, they are not high accuracy because traditional ones do not consider spatial characteristics. In this paper 'Pre-Filtering based Post Load Shedding' are suggested to improve the accuracy of spatial query and to restrict load shedding in DSMS. This method, at first, limits unnecessarily increased loads in stream queue with 'Pre-Filtering'. And then, it processes 'Post-Load Shedding', considering data and spatial status to guarantee the accuracy of result. The suggested method effectively reduces the number of the performance of load shedding, and improves the accuracy of spatial query.

• The Korean Journal of Applied Statistics
• /
• v.33 no.5
• /
• pp.615-625
• /
• 2020

The document-term frequency matrix is a term extracted from documents in which the group information exists in text mining. In this study, we generated the document-term frequency matrix for document classification according to research field. We applied the traditional term weighting function term frequency-inverse document frequency (TF-IDF) to the generated document-term frequency matrix. In addition, we applied term frequency-inverse gravity moment (TF-IGM). We also generated a document-keyword weighted matrix by extracting keywords to improve the document classification accuracy. Based on the keywords matrix extracted, we classify documents using a deep neural network. In order to find the optimal model in the deep neural network, the accuracy of document classification was verified by changing the number of hidden layers and hidden nodes. Consequently, the model with eight hidden layers showed the highest accuracy and all TF-IGM document classification accuracy (according to parameter changes) were higher than TF-IDF. In addition, the deep neural network was confirmed to have better accuracy than the support vector machine. Therefore, we propose a method to apply TF-IGM and a deep neural network in the document classification.

• The Journal of the Korea Contents Association
• /
• v.10 no.10
• /
• pp.54-67
• /
• 2010

Applying the conventional machine-learning method which has been frequently used in health-care area has several fundamental problems for modern U-health service analysis. First of all, we are still lack of application examples of the traditional method for our modern U-health environment because of its short term history of U-health study. Second, it is difficult to apply the machine-learning method to our U-health service environment which requires real-time management of disease because the method spends a lot of time in the process of learning. Third, we cannot implement a personalized U-health diagnosis system using the conventional method because there is no way to assign weights on the disease-related variables although various kinds of machine-learning schemes have been proposed. In this paper, a novel diagnosis scheme PCADP is proposed to overcome the problems mentioned above. PCADP scheme is a personalized diagnosis method and it makes the bio-data analysis just a 'process' in the U-health service system. In addition, we offer a semantics modeling of the U-health ontology framework in order to describe U-health data and service specifications as meaningful representations based on this PCADP. The PCADP scheme is a kind of statistical diagnosis method which has characteristics of flexible structure, real-time processing, continuous improvement, and easy monitoring of decision process. Upto the best of authors' knowledge, the PCADP scheme and ontology framework proposed in this paper reveals one of the best characteristics of flexible structure, real-time processing, continuous improvement, and easy monitoring among recently developed U-health schemes.

• Korea Journal of Hospital Management
• /
• v.6 no.2
• /
• pp.37-69
• /
• 2001

A new cost management system, called Activity Based Costing (ABC) system, has arisen to solve the limitation of a Traditional Cost Accounting (TCA) system until last two decades and ABC has been applied by many companies. TCA systems have limitation in tracing cost because they arbitrarily allocate overhead cost to the cost objects without standard for direct cost distribution. ABC is an accounting system that assigns costs to products or services based on the resources they consume. The costs of all activities are traced to the products for which they are performed. Therefore ABC is a cost management system that provides a matrix to accurately quantify consumed resources triggered by activities and activities triggered by products and services. There is little implementation of ABC in the health services field, one of service industries, due to complicated and many activities, and volatile cost object. However, the necessity for applying reasonable cost accounting system is largely issuing as strategy responding hostile environment, and financial pressure, and it is imperative to implement the Activity Based Costing (ABC) system. Therefore, this study presents the framework to develop ABC system for total health service organizations. Cost objects in this study base on medical service activities per health insurance claim from one general hospital located in Metropolitan Statistical Areas (MSAs). Medical service activities include all health insurance claims in the hospital. The purpose of the study is presenting useful tools and basic frame to develop Activity Based Costing system for health service organizations which want to use ABC system. The steps to develop ABC system for health service organizations are following: 1. Identifying of activity centers; 2. Definition of cost objects and activity by activity center; 3. Analysis of activity and tracing activity contribution; 4. Allocation of direct cost for specific activity; 5. Allocation of indirect cost for specific activity; 6. Allocation of depreciation for facilities, applicants, and consumption goods; 7. Allocation of administration cost; 8. Allocation of cost among activity centers; and 9. Tracing cost of cost objects by activity center. This study identified necessary information from existing reports which hospitals generally made by each step, and defined outcome which had to be produced in each step using this information. The steps of this study had limitation to apply all different size hospitals because the steps were structured ABC system by one hospital, however, this study used similar basic framework and methods with general cases. When a health service organization want to apply Activity Based Costing (ABC) system on all activities of it in future days, this study is very useful to design system structure in the health service organization.

• Journal of Periodontal and Implant Science
• /
• v.39 no.2
• /
• pp.111-118
• /
• 2009

Purpose: The integrity of interproximal hard/soft tissue has been widely accepted as the key determinant for success or degree of root coverage following the connective tissue graft. However, we reason that the gingival biotype of an individual, defined as the distance from the interproximal papilla to gingiva margin, may be the key determinant that influence the extent of root coverage regardless of traditional classification of gingival recession. Hence, the present study was performed with an aim to verify that individual gingival scalloping pattern inherent from biotype influence the level of gingival margin following the connective tissue graft for root coverage. Methods: Test group consisted of 43 single-rooted teeth from 21 patients (5 male and 16 female patients, mean age: 36.6 years) with varying degrees of gingival recession requiring connective tissue graft; 20 teeth of Miller class I and 23 teeth of Miller class III gingival recession, respectively. The control group consisted of contralateral teeth which did not demonstrate apparent gingival recession, and thus not requiring root coverage. For a biotype determination, an imaginary line connecting two adjacent papillae of a test tooth was drawn. The distance from this line to gingival margin at mid-buccal point and this distance (P-M distance) was designated as "gingival biotype" for a given individual. The distance was measured at baseline and 3 to 6 months examinations postoperatively both in test and control groups. The differences in the distance between Miller class I and III were subject to statistical analysis by using Student.s t-test while those between the test and control groups within a given patient were by using paired t-test. Results: The P-M distance at 3 to 6 months postoperatively was not significantly different between Miller class I and Miller class III. It was not significantly different between the test and control group in a given patient, either, both in Miller class I and III. Conclusions: The amount of root coverage following the connective tissue graft was not dependent on Miller's classification, but rather was dependent on P-M distance, strongly implying that the gingival biotype of a given patient may play a critical impact on the level of gingival margin following connective tissue graft.

• Journal of Digital Convergence
• /
• v.15 no.9
• /
• pp.65-78
• /
• 2017

This study aims to analyze whether both childcare subsidy and childcare leave policies have moderating effects on the relationship between Korean women's value of children and their intentions for subsequent childbirth. The data are used from the 2015 National Survey on Fertility and Family Health and Welfare. Both hierarchical multiple regression analysis and moderated regression analysis are used for statistical analysis. The findings from the study are as follow. First, the ideal number of children, instrumental values, and emotional values have positive effects on the intentions for subsequent childbirth after controlling for background variables(level of education, income, age, and number of children). Second, childcare leave policy has no significant effect on the intentions for subsequent childbirth while childcare subsidy policy has the negative effect. Third, only childcare subsidy policy has moderating effects on the relationship between instrumental, emotional values, and the intentions for subsequent childbirth. These results suggest that policies enhancing the value of children should be implemented in addition to traditional birth rate policies. Furthermore, new birth rate polices are needed for those married women who have a high possibility of subsequent childbirth.

• Korean Journal of Food Science and Technology
• /
• v.28 no.6
• /
• pp.1171-1176
• /
• 1996

The changes in volatiles of the model system were analyzed by GC and GC-MS before and after retorting. The GC data were analyzed statistically by applying the analysis of variance, and 42 peaks were selected at 5% significance level. Multivariate statistical analysis was performed with these 42 peaks as independent variables. Through the stepwise discriminant analysis, 8 peaks, which corresponded to the compounds such as 2-heptanone, cis-3-hexenal, 2-pentyl-furan, 1-methyl-trans-1,2-cyclohexanediol, 2-hexanone, 3-octanone, trans, trans-nona-2,4-dienal and 1-octen-3-ol, were obtained in sequence to distinguish the samples with and without retorting. The principal component analysis of a set of 8 independent variables resulted in 3 principal components which accounted for 96.1% of the variance, while the first principal component (PC 1) explained 76.5% of the total variance. In addition, through the factor analysis of the principal components, the peaks 11, 20 and 21 could be grouped togather in accordance with the direction and the size while the peaks 9, 33 and 39 constituted the second group in the direction.

• Land and Housing Review
• /
• v.4 no.4
• /
• pp.317-332
• /
• 2013

The purpose of this study is to survey the community support program of Green Belt from 2001 to 2011 and propose the improvement of the institution. For research method, the projects were analyzed by year, area, and category using statistical data. The improvement of the institution was drawn through the opinion survey of the interest group such as residents and public servants. For 10 years, 2007 community support projects were carried out and the total amount of government expenditure was 583.9 billion won. Among the support items, life convenience projects comprise 96.7%. For area, metropolitan area comprise 32.5%. There is a bias in items and areas. According to the survey of the residents and public servants, the satisfaction for the community support program is increasing. But it is necessary to enlarge the direct life cost support, activate community involvement and develop new project type. Proposed new projects are such as making characteristic village for income creation, planning for the landscape preservation using historic and traditional resources, making leisure space for nearby residents, and projects for the aged people. And it is proposed to give incentive to the characteristic village projects through competition.