• Clean Technology
• /
• 제26권4호
• /
• pp.329-335
• /
• 2020

Methanol synthesized from renewable hydrogen and captured CO2 has recently attracted great interest as a sustainable energy carrier for large-scale renewable energy storage. In this study, molten carbonate fuel cell's performance was investigated with the direct conversion of methanol into syngas inside the anode chamber of the cell. The internal reforming of methanol may significantly improve system efficiency since the heat generated from the electrochemical reaction can be used directly for the endothermic reforming reaction. The porous Ni-10 wt%Cr anode was sufficient for the methanol steam reforming reaction under the fuel cell operating condition. The direct supply of methanol into the anode chamber resulted in somewhat lower cell performance, especially at high current density. Recycling of the product gas into the anode gas inlet significantly improved the cell performance. The analysis based on material balance revealed that, with increasing current density and gas recycling ratio, the methanol steam reforming reaction rate likewise increased. A methanol conversion more significant than 90% was achieved with gas recycling. The results showed the feasibility of electricity and syngas co-production using the molten carbonate fuel cell. Further research is needed to optimize the fuel cell operating conditions for simultaneous production of electricity and syngas, considering both material and energy balances in the fuel cell.

• Journal of Korea Entertainment Industry Association
• /
• 제13권4호
• /
• pp.1-15
• /
• 2019

Psy's Gangnam style in 2012 showed K-pop's potential for global growth and BTS proved it by reaching three consecutive Billboard No.1. The success in the global music market brings tremendous economical and cultural power. This study is conducted for the continuous growth of K-pop music in the global music market by analyzing the musical factor of K-pop's global success. The top 20 most-viewed K-pop MV on Youtube is chosen as a research subject because Youtube is a worldwide platform that reflects global popularity. For the process of K-pop music creation, the role of the composer is expanded and many overseas producers participate in music creation. All 20 songs are created by the collective creation system and there is a consecutive collaboration between the main producers and certain artists. The top 20 most viewed K-pop songs have the musical characteristics of transnational genre convergence, hook songs, sophisticated sounds, frequent use of English lyrics, a reflection of the latest global trends, rhythm optimized for dance and clear concept. It makes the K-pop song easily remembered and familiar to overseas listeners. K-pop's healthy and fresh theme brings emotional empathy and reflects Korean sentiments. K-pop's global success is not a coincidence, but a result of continuous efforts to advance overseas. Some critics criticize K-pop's musical style is similar and it shows K-pop's limitation but K-pop progressed its musical evolution. By keeping the merits of K-pop's success factors and complementing its weak points, K-pop will continue its popularity and increase influence in the global music market.

• Journal of the Daesoon Academy of Sciences
• /
• 제30집
• /
• pp.1-57
• /
• 2018

There are three purposes to this study: first, to understand comprehensively the meaning of Haewon-sangsaeng (Resolution of Grievances for Mutual Beneficence) Thought, which can be taken as representative thought regarding peace in Korean new religions. Next, Haewon-sangsaeng Thought and the works for Haewon (resolving grievances) will be examined as principles and practical mechanisms for building the paradise of the Later World and understanding the structure of this system of thought. Lastly, logical inferences will be made regarding the future of humanity and the world through the ideological characteristics implied by Haewon-sangsaeng Thought. Haewon-sangsaeng Thought contains the complicated concepts of Haewon and Sangsaeng. Haewon is the resolution of the enmity and grievances that have accumulated in the realms of humanity and deities. Sangsaeng indicates the action of mutually benefiting one another or a state wherein people live in prosperity and peace. In Daesoon Jinrihoe, the concept of Haewon-sangsaeng is expressed explicitly and has broad applications. It can be expanded for the global peace and the harmony of all humanity. As the result of an integrated analysis of previous studies, it can be stated that Haewon-sangsaeng has values and meanings in terms of principles, laws, ethics, and ideology all of which are commonly connected to Injon (Human Nobility), Sangsaeng, peace, harmony, the Later World, and paradise. This indicates that its valuable for the future of humanity and world is deeper and wider than its mere etymological meaning. The common factor among paired ideas such as human nobility and Sangsaeng, peace and harmony, and Later World and paradise is the realization of humanity's greatest wish. This is the reason why the value and meaning of Haewon-sangsaeng can be expanded globally. The works of Haewon were a religious act of Kang Jeungsan who resolved the grievances of the Former World which was under the rule of mutual conflict and built a Later World that will operate according to mutual beneficence. Therefore, the principle of Haewon-sangsaeng has a motivative power, through the Reordering Works of the Universe, which can transform the future of humanity and the world. In this study, it can be inferred that as Haewon-sangsaeng 'fulfills human desires' and forms a 'harmonious relations of Sangsaeng' between humans and world, humans will be transformed into Injon (Human Nobility) while the world turns into a paradise, and the future turns into period of peace. Therefore, Haewon-sangsaeng Thought works as a principle that changes society, the world, and the universe. The social actualization of Haewon-sangsaeng is tantamount to bringing the future of Injon, paradise, and peace into objective reality. Previous studies on Haewon-sangsaeng Thought had been carried out under difficult circumstances by a small number of scholars. For all the above reasons, I anticipate that there will be more and more studies made on the topic of Haewon-sangsaeng Thought, which seeks the realization of Haewon (the Resolution of Grievances), Sangsaeng (Mutual Beneficence), human nobility, paradise, and peace. I hope it will emerge as a main subject in global religious thought.

• Proceedings of the Korea Water Resources Association Conference
• /
• 한국수자원학회 2022년도 학술발표회
• /
• pp.216-217
• /
• 2022

Gulpur 수력발전 프로젝트는 전력난을 겪고 있는 파키스탄에 102 MW 규모의 수력발전소를 건설하여 30년 동안 운영 관리한 후 파키스탄 정부로 양도하는 IPP(Independent Power Producing) 형식의 투자사업이다. 남동발전과 DL E&C, 롯데건설이 Sponsor로서 출자한 자본금과, ADB, IFC, K-EXIM 등의 대주단로부터의 차입금을 재원으로 하여 소요 사업비를 조달하고 사업을 개발하였다. DL E&C와 롯데건설이 EPC(Engineering, Procurement, Construction)를 수행하였고, 이산이 Design consultant의 역할을 수행하였다. Gulpur 수력발전 프로젝트의 발전형식은 수로식(run-of-river)으로 201 m³/s의 발전유량과 102 MW의 발전 시설용량을 이용하여 연평균예상발전량은 398 GWh이다. 주요 구조물로는 설계 재현빈도 1년의 유수전환시설(가물막이댐 & 가배수터널)과 콘크리트 중력식댐(H 67 m, L 205 m), 도수터널(D 6.7 m, L 215 m, 2기), 옥외형 발전소 (H 51 m, W 60 m, L 38 m, Kaplan 2기)가 있으며, 2015년 10월 착공하여 2020년 3월 상업발전을 시작하였다. 본 프로젝트는 DL E&C의 첫 번째 EPC 해외수력발전 프로젝트이다. 따라서 프로젝트의 성공적 수행을 위한 경제적 설계, 시공의 효율성 및 안정성 확보 등을 위하여 많은 연구를 수행하는 과정에서 다양한 기술 개선을 이룰 수 있었다. 본고에서는 Gulpur 프로젝트를 통하여 도출된 성공 사례들을 소개 및 공유하고자 한다. 첫 번째로 콘크리트 중력식댐 시공을 위한 유수전환시설의 최적 설계빈도를 산정하였다. 일반적으로 유수전환시설의 규모는 설계기준에 제시된 설계 재현빈도를 이용하는데, 해외 설계기준에서는 10년, 국내 설계기준에서는 1~2년으로 다르게 제시되어 있는 문제점이 있다. 유수전환시설의 규모는 프로젝트의 경제성에 큰 영향을 미치기 때문에 최적 설계빈도의 결정이 필요하며, 위험도분석기법(Risk Analysis)과 기대화폐가치법(Expected Monetary Value)을 이용하여 유수전환시설의 최적 설계 재현빈도와 이에 영향을 미치는 인자를 분석하였다. 위험도는 몬테카를로 시뮬레이션으로 산정된 가물막이댐 파괴확률과 재현빈도를 이용하여 산정된 가물막이댐 월류확률을 고려하였으며, 비용 및 피해액으로는 유수전환시설의 공사비, 가물막이댐 파괴시의 재건설비용과 지체보상금, 가물막이댐 월류시의 복구비용을 고려하였다. 이에 대한 연구결과로, 유수전환시설의 사용기간과 월류시의 복구비용이 유수전환시설의 설계 재현기간 결정에 가장 큰 영향을 미치는 것으로 나타났고, 특히 월류시의 복구비용이 작을수록 낮은 설계 재현빈도를 선택하는 것이 타당한 것으로 나타났다. 예를 들어, 유수전환시설의 사용기간이 3 ~ 5년, 복구비용이 0.5 ~ 1.0 mil USD 이하인 조건에서 가물막이시설의 최적 설계빈도는 1년 ~ 2년인 것으로 나타났다. 또한, 유수전환시설의 사용기간은 본댐의 규모와 시공기간 등을 고려하여 결정되는 사항으로 설계자가 임의 조정할 수 없지만, 복구비용은 시공 관리자에 따라 결정되는 부분으로, 적극적 홍수 피해 저감 및 복구방안을 마련하는 것이 프로젝트의 경제성을 향상시킬 수 있다는 것을 알 수 있었다. 두 번째로 프로젝트의 경제성 향상, 홍수기 댐 시공시의 안전성 확보를 위하여 홍수 조기경보시스템(Early Warning System)을 개발 및 활용하였다. 수로식(Run-of-river) 수력발전댐은 대부분 산악지역에 위치하기 때문에 국지성 강우 및 급한 지형 경사로 인하여 돌발홍수(flash flood)의 발생 가능성이 높다. 따라서 시공 중 홍수(월류) 발생을 미리 감지하고 현장에 전파할 수 있는, 수로식(Run-of-river) 수력발전댐 현장을 위한 홍수 조기경보시스템이 필요하며, 이를 리스크 인식, 모니터링 및 경보, 전파 및 연락, 반응 능력 향상의 4가지 부분으로 나누어 구축하였다. 리스크 인식 부분에서는 가물막이댐 월류 발생 상황에 대한 위험도, 취약성, 리스크를 제시하였으며, 모니터링 및 경보 부분에서는 상류 측정수위에서 유도된 현장 예상수위와 실제 현장 측정 수위를 대상으로 경보홍수위와 위험홍수위로 나누어 관리하였다. 전파 및 연락 부분에서는 현장 시공 조직을 활용하여 홍수시를 대비한 비상연락체계도(Emergency communication flow chart)를 운영하였으며, 반응 능력 향상을 위해 비상연락체계도의 팀별 Action plan을 상세화 하였다. 세 번째로 현장의 지질특성과 50여 차례 발파시험으로 현장 고유의 발파진동감쇄곡선을 도출하였으며, 이를 통해 현장의 시공성과 콘크리트 품질 확보를 동시에 달성할 수 있는 방안을 제시하였다. 콘크리트댐 공사에서는 제한된 공기 내에 공사를 완료하기 위해 사면부 굴착과 콘크리트 타설이 동시에 수행될 수밖에 없는 문제점을 가지고 있다. 그러나 신규 콘크리트 타설면 근처에서 발파를 수행하는 경우 발파로 발생되는 탄성파가 일정 수준을 초과하게 되면, 콘크리트 양생에 영향을 주게 된다. 따라서 다수의 현장 발파시험을 통해 발파거리와 최대진동속도의 상관관계 즉, 발파진동감쇄곡선을 도출함으로써 현장의 발파진동특성을 도출할 수 있었다. 또한, 기존 연구 논문들을 통해 콘크리트 재령기간 별 안전진동속도를 선정하고, 해당 안전진동속도를 초과하지 않는 범위에서 콘크리트 타설면과 발파위치의 거리에 따라 1회 발파 가능한 장약량을 산정하여 적용하였다. 이와 같은 체계적인 접근을 통해 콘크리트 타설과 발파 작업 동시 수행에 대한 논란을 해소할 수 있었다.

• Asia-Pacific Journal of Business Venturing and Entrepreneurship
• /
• 제17권2호
• /
• pp.33-51
• /
• 2022

This study analyzes newly listed companies on KOSDAQ from 2011 to 2020 for both firms having experience in attracting venture investment before listing (VI) and those without having experience in attracting venture investment (NVI) by examining differences between two groups (VI and NVI) with respect to both the level of listing performance and that of firm performance (growth) after the listing. This paper conducts descriptive statistics, mean difference, and multiple regression analysis. Independent variables for regression models include VC investment, firm age at the time of listing, firm type, firm location, firm size, the age of VC, the level of expertise of VC, and the level of fitness of VC with investment company. Throughout this paper, results suggest that listing performance and post-listed growth are better for VI than NVI. VC investment shows a negative effect on the listing period and a positive effect on the sales growth rate. Also, the amount of VC investment has negative effects on the listing period and positive effects on the market capitalization at the time of IPO and on sales growth among growth indicators. Our evidence also implies a significantly positive effect on growth after listing for firms which belong to R&D specialized industries. In addition, it is statistically significant for several years that the firm age has a positive effect on the market capitalization growth rate. This shows that market seems to put the utmost importance on a long-term stability of management capability. Finally, among the VC characteristics such as the age of VC, the level of expertise of VC, and the level of fitness of VC with investment company, we point out that a higher market capitalization tends to be observed at the time of IPO when the level of expertise of anchor VC is high. Our paper differs from prior research in that we reexamine the venture ecosystem under the outbreak of coronavirus disease 2019 which stimulates the degradation of the business environment. In addition, we introduce more effective variables such as VC investment amount when examining the effect of firm type. It enables us to indirectly evaluate the validity of technology exception policy. Although our findings suggest that related policies such as the technology special listing system or the injection of funds into the venture ecosystem are still helpful, those related systems should be updated in a more timely fashion in order to support growth power of firms due to the rapid technological development. Furthermore, industry specialization is essential to achieve regional development, and the growth of the recovery market is also urgent.

• Journal of Intelligence and Information Systems
• /
• 제19권4호
• /
• pp.123-132
• /
• 2013

As the smartphones are equipped with various sensors such as the accelerometer, GPS, gravity sensor, gyros, ambient light sensor, proximity sensor, and so on, there have been many research works on making use of these sensors to create valuable applications. Human activity recognition is one such application that is motivated by various welfare applications such as the support for the elderly, measurement of calorie consumption, analysis of lifestyles, analysis of exercise patterns, and so on. One of the challenges faced when using the smartphone sensors for activity recognition is that the number of sensors used should be minimized to save the battery power. When the number of sensors used are restricted, it is difficult to realize a highly accurate activity recognizer or a classifier because it is hard to distinguish between subtly different activities relying on only limited information. The difficulty gets especially severe when the number of different activity classes to be distinguished is very large. In this paper, we show that a fairly accurate classifier can be built that can distinguish ten different activities by using only a single sensor data, i.e., the smartphone accelerometer data. The approach that we take to dealing with this ten-class problem is to use the ensemble of nested dichotomy (END) method that transforms a multi-class problem into multiple two-class problems. END builds a committee of binary classifiers in a nested fashion using a binary tree. At the root of the binary tree, the set of all the classes are split into two subsets of classes by using a binary classifier. At a child node of the tree, a subset of classes is again split into two smaller subsets by using another binary classifier. Continuing in this way, we can obtain a binary tree where each leaf node contains a single class. This binary tree can be viewed as a nested dichotomy that can make multi-class predictions. Depending on how a set of classes are split into two subsets at each node, the final tree that we obtain can be different. Since there can be some classes that are correlated, a particular tree may perform better than the others. However, we can hardly identify the best tree without deep domain knowledge. The END method copes with this problem by building multiple dichotomy trees randomly during learning, and then combining the predictions made by each tree during classification. The END method is generally known to perform well even when the base learner is unable to model complex decision boundaries As the base classifier at each node of the dichotomy, we have used another ensemble classifier called the random forest. A random forest is built by repeatedly generating a decision tree each time with a different random subset of features using a bootstrap sample. By combining bagging with random feature subset selection, a random forest enjoys the advantage of having more diverse ensemble members than a simple bagging. As an overall result, our ensemble of nested dichotomy can actually be seen as a committee of committees of decision trees that can deal with a multi-class problem with high accuracy. The ten classes of activities that we distinguish in this paper are 'Sitting', 'Standing', 'Walking', 'Running', 'Walking Uphill', 'Walking Downhill', 'Running Uphill', 'Running Downhill', 'Falling', and 'Hobbling'. The features used for classifying these activities include not only the magnitude of acceleration vector at each time point but also the maximum, the minimum, and the standard deviation of vector magnitude within a time window of the last 2 seconds, etc. For experiments to compare the performance of END with those of other methods, the accelerometer data has been collected at every 0.1 second for 2 minutes for each activity from 5 volunteers. Among these 5,900 ($=5{\times}(60{\times}2-2)/0.1$) data collected for each activity (the data for the first 2 seconds are trashed because they do not have time window data), 4,700 have been used for training and the rest for testing. Although 'Walking Uphill' is often confused with some other similar activities, END has been found to classify all of the ten activities with a fairly high accuracy of 98.4%. On the other hand, the accuracies achieved by a decision tree, a k-nearest neighbor, and a one-versus-rest support vector machine have been observed as 97.6%, 96.5%, and 97.6%, respectively.

• Journal of Intelligence and Information Systems
• /
• 제24권1호
• /
• pp.205-225
• /
• 2018

Convolutional Neural Network (ConvNet) is one class of the powerful Deep Neural Network that can analyze and learn hierarchies of visual features. Originally, first neural network (Neocognitron) was introduced in the 80s. At that time, the neural network was not broadly used in both industry and academic field by cause of large-scale dataset shortage and low computational power. However, after a few decades later in 2012, Krizhevsky made a breakthrough on ILSVRC-12 visual recognition competition using Convolutional Neural Network. That breakthrough revived people interest in the neural network. The success of Convolutional Neural Network is achieved with two main factors. First of them is the emergence of advanced hardware (GPUs) for sufficient parallel computation. Second is the availability of large-scale datasets such as ImageNet (ILSVRC) dataset for training. Unfortunately, many new domains are bottlenecked by these factors. For most domains, it is difficult and requires lots of effort to gather large-scale dataset to train a ConvNet. Moreover, even if we have a large-scale dataset, training ConvNet from scratch is required expensive resource and time-consuming. These two obstacles can be solved by using transfer learning. Transfer learning is a method for transferring the knowledge from a source domain to new domain. There are two major Transfer learning cases. First one is ConvNet as fixed feature extractor, and the second one is Fine-tune the ConvNet on a new dataset. In the first case, using pre-trained ConvNet (such as on ImageNet) to compute feed-forward activations of the image into the ConvNet and extract activation features from specific layers. In the second case, replacing and retraining the ConvNet classifier on the new dataset, then fine-tune the weights of the pre-trained network with the backpropagation. In this paper, we focus on using multiple ConvNet layers as a fixed feature extractor only. However, applying features with high dimensional complexity that is directly extracted from multiple ConvNet layers is still a challenging problem. We observe that features extracted from multiple ConvNet layers address the different characteristics of the image which means better representation could be obtained by finding the optimal combination of multiple ConvNet layers. Based on that observation, we propose to employ multiple ConvNet layer representations for transfer learning instead of a single ConvNet layer representation. Overall, our primary pipeline has three steps. Firstly, images from target task are given as input to ConvNet, then that image will be feed-forwarded into pre-trained AlexNet, and the activation features from three fully connected convolutional layers are extracted. Secondly, activation features of three ConvNet layers are concatenated to obtain multiple ConvNet layers representation because it will gain more information about an image. When three fully connected layer features concatenated, the occurring image representation would have 9192 (4096+4096+1000) dimension features. However, features extracted from multiple ConvNet layers are redundant and noisy since they are extracted from the same ConvNet. Thus, a third step, we will use Principal Component Analysis (PCA) to select salient features before the training phase. When salient features are obtained, the classifier can classify image more accurately, and the performance of transfer learning can be improved. To evaluate proposed method, experiments are conducted in three standard datasets (Caltech-256, VOC07, and SUN397) to compare multiple ConvNet layer representations against single ConvNet layer representation by using PCA for feature selection and dimension reduction. Our experiments demonstrated the importance of feature selection for multiple ConvNet layer representation. Moreover, our proposed approach achieved 75.6% accuracy compared to 73.9% accuracy achieved by FC7 layer on the Caltech-256 dataset, 73.1% accuracy compared to 69.2% accuracy achieved by FC8 layer on the VOC07 dataset, 52.2% accuracy compared to 48.7% accuracy achieved by FC7 layer on the SUN397 dataset. We also showed that our proposed approach achieved superior performance, 2.8%, 2.1% and 3.1% accuracy improvement on Caltech-256, VOC07, and SUN397 dataset respectively compare to existing work.

• Journal of Nutrition and Health
• /
• 제48권2호
• /
• pp.180-191
• /
• 2015

Purpose: There is a lack of data on the energy cost of children's everyday activities, adult values are often used as surrogates. In addition, the influence of body weight on the energy cost of activity when expressed as metabolic equivalents (METs) has not been vigorously explored. Methods: In this study 20 elementary school students 9~12 years of age completed 18 various physical activities while energy expenditure was measured continuously using a portable telemetry gas exchange system ($K_4b^2$, Cosmed, Rome, Italy). Results: The average age was 10.4 years and the average height and weight was 145.1 cm and 43.6 kg, respectively. Oxygen consumption ($VO_2$), energy expenditure and METs at the time of resting of the subjects were 5.41 mL/kg/min, 1.44 kcal/kg/h, and 1.5 METs, respectively. METs values by 18 physical activities were as follows: Homework and reading books (1.6 METs), playing game with a mobile phone or video while sitting (1.6 METs), watching TV while sitting on a comfortable chair (1.7 METs), playing video game or mobile phone game while standing (1.9 METs), sweeping a room with a broom (2.7 METs) and playing a board game (2.8 METs) belong to light intensity physical activities. By contrary, speedy walking and running were 6.6 and 6.7 METs, respectively, which belong to high intensity physical activities over 6.0 METs. When the effect of body weight on physical activity energy expenditure was determined, $R^2$ values increased with 0.116 (playing a game at sitting), 0.176 (climbing up and down stairs), 0.246 (slow walking), and 0.455 (running), which showed that higher activity intensity increased explanation power of body weight on METs value. Conclusion: This study is important for direct evaluation of energy expenditure by physical activities of children, and it could be used directly for revising and complementing the existing activity classification table to fit for children.

• Journal of Intelligence and Information Systems
• /
• 제24권2호
• /
• pp.111-124
• /
• 2018

This study aims to develop a classification model for predicting the occurrence of hyperlipidemia, one of the chronic diseases. Prior studies applying data mining techniques for predicting disease can be classified into a model design study for predicting cardiovascular disease and a study comparing disease prediction research results. In the case of foreign literatures, studies predicting cardiovascular disease were predominant in predicting disease using data mining techniques. Although domestic studies were not much different from those of foreign countries, studies focusing on hypertension and diabetes were mainly conducted. Since hypertension and diabetes as well as chronic diseases, hyperlipidemia, are also of high importance, this study selected hyperlipidemia as the disease to be analyzed. We also developed a model for predicting hyperlipidemia using SVM and meta learning algorithms, which are already known to have excellent predictive power. In order to achieve the purpose of this study, we used data set from Korea Health Panel 2012. The Korean Health Panel produces basic data on the level of health expenditure, health level and health behavior, and has conducted an annual survey since 2008. In this study, 1,088 patients with hyperlipidemia were randomly selected from the hospitalized, outpatient, emergency, and chronic disease data of the Korean Health Panel in 2012, and 1,088 nonpatients were also randomly extracted. A total of 2,176 people were selected for the study. Three methods were used to select input variables for predicting hyperlipidemia. First, stepwise method was performed using logistic regression. Among the 17 variables, the categorical variables(except for length of smoking) are expressed as dummy variables, which are assumed to be separate variables on the basis of the reference group, and these variables were analyzed. Six variables (age, BMI, education level, marital status, smoking status, gender) excluding income level and smoking period were selected based on significance level 0.1. Second, C4.5 as a decision tree algorithm is used. The significant input variables were age, smoking status, and education level. Finally, C4.5 as a decision tree algorithm is used. In SVM, the input variables selected by genetic algorithms consisted of 6 variables such as age, marital status, education level, economic activity, smoking period, and physical activity status, and the input variables selected by genetic algorithms in artificial neural network consist of 3 variables such as age, marital status, and education level. Based on the selected parameters, we compared SVM, meta learning algorithm and other prediction models for hyperlipidemia patients, and compared the classification performances using TP rate and precision. The main results of the analysis are as follows. First, the accuracy of the SVM was 88.4% and the accuracy of the artificial neural network was 86.7%. Second, the accuracy of classification models using the selected input variables through stepwise method was slightly higher than that of classification models using the whole variables. Third, the precision of artificial neural network was higher than that of SVM when only three variables as input variables were selected by decision trees. As a result of classification models based on the input variables selected through the genetic algorithm, classification accuracy of SVM was 88.5% and that of artificial neural network was 87.9%. Finally, this study indicated that stacking as the meta learning algorithm proposed in this study, has the best performance when it uses the predicted outputs of SVM and MLP as input variables of SVM, which is a meta classifier. The purpose of this study was to predict hyperlipidemia, one of the representative chronic diseases. To do this, we used SVM and meta-learning algorithms, which is known to have high accuracy. As a result, the accuracy of classification of hyperlipidemia in the stacking as a meta learner was higher than other meta-learning algorithms. However, the predictive performance of the meta-learning algorithm proposed in this study is the same as that of SVM with the best performance (88.6%) among the single models. The limitations of this study are as follows. First, various variable selection methods were tried, but most variables used in the study were categorical dummy variables. In the case with a large number of categorical variables, the results may be different if continuous variables are used because the model can be better suited to categorical variables such as decision trees than general models such as neural networks. Despite these limitations, this study has significance in predicting hyperlipidemia with hybrid models such as met learning algorithms which have not been studied previously. It can be said that the result of improving the model accuracy by applying various variable selection techniques is meaningful. In addition, it is expected that our proposed model will be effective for the prevention and management of hyperlipidemia.

• Journal of Intelligence and Information Systems
• /
• 제24권2호
• /
• pp.221-241
• /
• 2018

Deep learning is getting attention recently. The deep learning technique which had been applied in competitions of the International Conference on Image Recognition Technology(ILSVR) and AlphaGo is Convolution Neural Network(CNN). CNN is characterized in that the input image is divided into small sections to recognize the partial features and combine them to recognize as a whole. Deep learning technologies are expected to bring a lot of changes in our lives, but until now, its applications have been limited to image recognition and natural language processing. The use of deep learning techniques for business problems is still an early research stage. If their performance is proved, they can be applied to traditional business problems such as future marketing response prediction, fraud transaction detection, bankruptcy prediction, and so on. So, it is a very meaningful experiment to diagnose the possibility of solving business problems using deep learning technologies based on the case of online shopping companies which have big data, are relatively easy to identify customer behavior and has high utilization values. Especially, in online shopping companies, the competition environment is rapidly changing and becoming more intense. Therefore, analysis of customer behavior for maximizing profit is becoming more and more important for online shopping companies. In this study, we propose 'CNN model of Heterogeneous Information Integration' using CNN as a way to improve the predictive power of customer behavior in online shopping enterprises. In order to propose a model that optimizes the performance, which is a model that learns from the convolution neural network of the multi-layer perceptron structure by combining structured and unstructured information, this model uses 'heterogeneous information integration', 'unstructured information vector conversion', 'multi-layer perceptron design', and evaluate the performance of each architecture, and confirm the proposed model based on the results. In addition, the target variables for predicting customer behavior are defined as six binary classification problems: re-purchaser, churn, frequent shopper, frequent refund shopper, high amount shopper, high discount shopper. In order to verify the usefulness of the proposed model, we conducted experiments using actual data of domestic specific online shopping company. This experiment uses actual transactions, customers, and VOC data of specific online shopping company in Korea. Data extraction criteria are defined for 47,947 customers who registered at least one VOC in January 2011 (1 month). The customer profiles of these customers, as well as a total of 19 months of trading data from September 2010 to March 2012, and VOCs posted for a month are used. The experiment of this study is divided into two stages. In the first step, we evaluate three architectures that affect the performance of the proposed model and select optimal parameters. We evaluate the performance with the proposed model. Experimental results show that the proposed model, which combines both structured and unstructured information, is superior compared to NBC(Naïve Bayes classification), SVM(Support vector machine), and ANN(Artificial neural network). Therefore, it is significant that the use of unstructured information contributes to predict customer behavior, and that CNN can be applied to solve business problems as well as image recognition and natural language processing problems. It can be confirmed through experiments that CNN is more effective in understanding and interpreting the meaning of context in text VOC data. And it is significant that the empirical research based on the actual data of the e-commerce company can extract very meaningful information from the VOC data written in the text format directly by the customer in the prediction of the customer behavior. Finally, through various experiments, it is possible to say that the proposed model provides useful information for the future research related to the parameter selection and its performance.