• Journal of Korean Institute of Industrial Engineers
• /
• v.25 no.1
• /
• pp.111-124
• /
• 1999

This evapaper is toluate the forecasting performance of three neural network(NN) approaches against ARIMA model using the famous time series analysis competition data. The first NN approach is to analyze the second Makridakis (M2) Competition Data using Multilayer Perceptron (MLP) that has been the most popular NN model in time series analysis. Since it is recently known that MLP suffers from bias/variance dilemma, two approaches are suggested in this study. The second approach adopts Cascade Correlation Network (CCN) that was suggested by Fahlman & Lebiere as an alternative to MLP. In the third approach, a time series is separated into two series using Noise Filtering Network (NFN) that utilizes autoassociative memory function of neural network. The forecasts in the decomposition analysis are the sum of two prediction values obtained from modeling each decomposed series, respectively. Among the three NN approaches, Decomposition Analysis shows the best forecasting performance on the M2 Competition Data, and is expected to be a promising tool in analyzing socio-economic time series data because it reduces the effect of noise or outliers that is an impediment to modeling the time series generating process.

• Proceedings of the Korean Society for Rock Mechanics Conference
• /
• 2000.09a
• /
• pp.129-137
• /
• 2000

This paper presents groundwater flow characteristics in discontinuous rock mass using fracture network program(NAPSAC) by statistical approach. Equivalent permeability coefficients are estimated from borehole data around Mabuk test tunnel site and fracture map on the arch of the tunnel. The reliability of fracture network model is obtained from determination of input data for statistical fracture network analysis from the real data(data of fracture network, data of hydraulic tests). The variation of permeability and mean anisotropic permeability coefficients are calculated from the realized model by increasing the size. As a result of analysis, a strong anisotropy of permeability is observed according to the direction of the fracture sets around the test tunnel.

• Tunnel and Underground Space
• /
• v.10 no.3
• /
• pp.378-386
• /
• 2000

This paper presents groundwater flow characteristics in discontinuous rock mass using fracture network program(NAPSAC) by statistical approach. Equivalent permeability coefficients are estimated from borehole data around Mabuk test tunnel site and fracture map on the arch of the tunnel. The reliability of fracture network model is obtained from determination of input data for statistical fracture network analysis from the real data(data of fracture network, data of hydraulic tests). The variation of permeability and mean anisotropic permeability coefficients are calculated from the realized model by increasing the size. As a result of analysis, a strong anisotropy of permeability is observed according to the direction of the fracture sets around the test tunnel.

• Korean Journal of Agricultural Science
• /
• v.47 no.3
• /
• pp.509-517
• /
• 2020

Geographic information system (GIS) sewer network data are a fundamental input material for urban inundation modeling, which is important to reduce the increasing damages from urban inundation due to climate change. However, the essential attributes of the data built by a local government are often missing because the purpose of building the data is the maintenance of the sewer system. Inconsistent simplification and supplementation of the sewer network data made by individual researchers may increase the uncertainty of flood simulations and influence the inundation analysis results. Therefore, it is necessary to develop a basic algorithm to convert the GIS-based sewage network data into input data that can be used for inundation simulations in consistent way. In this study, the format of GIS-based sewer network data for a watershed near the Sadang Station in Seoul and the Oncheon River Basin in Busan was investigated, and a missing data supplementing algorithm was developed. The missing data such as diameter, location, elevation of pipes and manholes were assumed following a consistent rule, which was developed referring to government documents, previous studies, and average data. The developed algorithm will contribute to minimizing the uncertainty of sewer network data in an urban inundation analysis by excluding the subjective judgment of individual researchers.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2021.07a
• /
• pp.117-122
• /
• 2021

In order to reveal the co-research relationship with key research institutes in the computer field in Korea, which have high competitiveness in the ICT field, this paper analyzes the co-author network of research institutes based on the papers of Korean journals. For this study, 15 computer journals published in Korea were selected and collected for the past 5 years. In addition, authors, affiliated institutions, and country information were extracted from the papers, and experimental data were produced using institutional identification data and SNA analysis techniques. Through this, statistical analysis was performed on researchers, research institutes, and countries, and the centrality of research institutes was measured through co-authored network analysis and visualized as a map. The results of the research that analyzed the co-author relationship based on the papers in computer science journals published in Korea are as follows. First, the status of co-author relations among researchers was statistically analyzed in detail. The number of authors, the number of co-authors, and statistics by co-authoring countries, as well as the ratio of co-authored research by institution and the ratio of co-authored by country were revealed. The co-authoring status of papers in English journals was analyzed separately to compare and analyze characteristics. Second, the centrality of each individual was measured through the co-authored network analysis of the identified researchers, institutions and countries affiliated with the institution, and the frequency of connection between data by level was identified. In addition, the co-authorization network between research institutes was visualized and presented as a map so that the overall view of the co-authoring network was presented, and the connection frequency between the institutes was checked. It is possible to recommend a research partner based on the co-authoring behavior of Korean computer researchers and the co-authoring relationship between major research institutes. For this, additional research is needed to analyze the subject areas of each institution.

• Journal of information and communication convergence engineering
• /
• v.14 no.1
• /
• pp.9-13
• /
• 2016

With the introduction of the data-oriented plan for LTE service, one may concerned with the background of the ISP's policy in charging for LTE services. In this work we investigate the latest usage patterns of voice and data applications for customers over the current mobile network, via which we investigate why mobile operators introduced data-oriented plans. To be specific, we collected the real-field data for the volume of voice and data traffic from the LTE network before the data-oriented plans were introduced. From the collected data we compute the absolute volume as well as the proportion of voice and data applications. From these observations we infer mobile operators' reasoning behind the decision to introduce data-oriented plans with unlimited voice calls over the mobile network.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.12
• /
• pp.4706-4724
• /
• 2020

With various structured data, such as the company size, loan balance, and savings accounts, the voice of customer (VOC), which is text data containing contact history and counseling details was analyzed in this study. To analyze unstructured data, the term frequency-inverse document frequency (TF-IDF) analysis, semantic network analysis, sentiment analysis, and a convolutional neural network (CNN) were implemented. A performance comparison of the models revealed that the predictive model using the CNN provided the best performance with regard to predictive power, followed by the model using the TF-IDF, and then the model using semantic network analysis. In particular, a character-level CNN and a word-level CNN were developed separately, and the character-level CNN exhibited better performance, according to an analysis for the Korean language. Moreover, a systematic selection model for optimal text mining techniques was proposed, suggesting which analytical technique is appropriate for analyzing text data depending on the context. This study also provides evidence that the results of previous studies, indicating that individual customers leave when their loyalty and switching cost are low, are also applicable to corporate customers and suggests that VOC data indicating customers' needs are very effective for predicting their behavior.

• 한국데이터정보과학회:학술대회논문집
• /
• 2003.10a
• /
• pp.57-72
• /
• 2003

This article provides integrated neural network models for the interest rate forecasting using change-point detection. The model is composed of three phases. The first phase is to detect successive structural changes in interest rate dataset. The second phase is to forecast change-point group with data mining classifiers. The final phase is to forecast the interest rate with BPN. Based on this structure, we propose three integrated neural network models in terms of data mining classifier: (1) multivariate discriminant analysis (MDA)-supported neural network model, (2) case based reasoning (CBR)-supported neural network model and (3) backpropagation neural networks (BPN)-supported neural network model. Subsequently, we compare these models with a neural network model alone and, in addition, determine which of three classifiers (MDA, CBR and BPN) can perform better. For interest rate forecasting, this study then examines the predictability of integrated neural network models to represent the structural change.

• The KIPS Transactions:PartA
• /
• v.17A no.4
• /
• pp.181-188
• /
• 2010

In this study, we suggest the system to analyze network users' action patterns by using Data-Mining Technique. We installed Network Tap to implement the analysis system of network action and copied the network packet. The copied packet is stored at the database through MainMemoryDB(MMDB) of the high-speed. The stored data analyze the users' action patterns by using Data-Mining Technique and then report the results to the network manager on real-time. Also, we applied the standard XML document exchange method to share the data between different systems. We propose this action pattern analysis system operated embedded type of SetToBox to install easily and support low price.

• Journal of Information Technology Services
• /
• v.10 no.1
• /
• pp.35-46
• /
• 2011

As development of exhibition, it has been an important issue to analyze exhibition for the next success of exhibition. A lot of existing researches have focused on the exhibitor's and visitor's satisfaction problem. However, the exhibitor's satisfaction and success of exhibition come from the analysis of exhibition in the network level. Booths composing exhibition are regarded as nodes in network, so the trace of visitors visiting booths can construct the arc of networks. The purpose of this study is to analyze the booth visiting pattern and components of network through social network analysis using data collected in the $17^{th}$ International KIDS & EDU EXPO for Children. This research is the first approach of network-leveled analysis of exhibition, and the result of network analysis is helpful to support the booth arrangement in next exhibition. Our analysis results the following implications. First, Booths with high degree centrality or betweenness centrality should be deployed in the wide space or corner of the exhibition hall. Finally, booths within a block should be deployed in the same space of the exhibition hall to provide convenience to the visitors and to enhance exhibition performance.