Category Archives: Machine Learning / Deep Learning

기계학습과 롤링 윈도우 기법을 활용한 주식시장 및 환율 예측 모델 구현

한재윤, 김진호, 황보유정, & 이경전. (2017). 기계학습과 롤링 윈도우 기법을 활용한 주식시장 및 환율 예측 모델 구현. 한국지능정보시스템학회 학술대회논문집, 69-70. (link)

ABSTRACT

최근 기계학습의 기법과 성능이 발전함에 따라, 금융권에서도 기계학습을 활용하여 주식시장 및 환율 등을 예측하려는 시도가 많아지고 있다. 하지만, 단순히 경제지표를 예측하는 경우, 변동성이 크다는 특징으로 인해 낮은 성능을 보이는 문제가 발생한다. 이에 본 연구에서는 주식 시장에 성질을 안녕 하는 기법들을 탐색, 활용하여 주식시장과 환율이 증감과 지수값을 높은 성능으로 예측하는 모델을 구현하였다. 다양한 국가 기반을 가진 18 개 지수를 사용하였으며, 변동성의 영향ㅇ을 최대한 줄이기 위해 롤링 윈도우 기법을 적용한 기계학습 모델을 구현하였다. 또한 타임래그와 로그 변환 등의 데이터 전처리 기법을 적용하여 기계학습 모델의 성능을 전반적으로 향상시켰다. 그 결과, 모든 주가지수의 증감에 대한 정확도는 평균 0.793 으로 높은 성능을 보였으며, 몇몇 변수에 대해서는 0.90 을 넘는 성능을 보였다. 또한 정화도, MAPE, RMSE, R2 등의 다양한 평가 기준에 대해서 더 좋은 성능을 보이는 타임래그가 존재한다는 사실을 확인하였다.

A Review on Key Research Topics of AI (Artificial Intelligence): promise and potential

Kwon, J. W., Kim, J. H., Park, A., & Lee, K. J. (2017). “A Review on Key Research Topics of AI (Artificial Intelligence): promise and potential”, In Proceedings of Korea Society of IT Service, 16(1), 511-514. (link) in Korean

ABSTRACT

현재 국내뿐만 아니라, 전 세계적으로 4차 산업혁명이 일어나고 있다. 4차 산업혁명의 중심에 있는 인공지능 기술은 그 종류와 적용 가능성이 매우 높고, 실제로 많은 산업에서 인공지능을 적용하여 비즈니스 목표를 달성하고 있다. 본 연구에서는 실제로 인공지능을 활용하여 산업에 적용한 논문들을 수집하여, 어떤 분야에서 어떠한 기술로 문제를 해결하고 있는지를 보고자 한다. 또한 비즈니스 관점에서 바라본 인공지능으로 달성할 수 있는 비즈니스 목표를 분석하였다.

Customer-driven smart and sustainable interaction in convention: Nestle’s IoT adoption case

Park, A., Jun, J. H., & Lee, K. J. (2017). Customer-driven smart and sustainable interaction in convention. 한국지능정보시스템학회 학술대회논문집, 134-135. (link)

ABSTRACT

Services based on the IoT technologies have emerged in various business environments. To enhance service quality for convention, maximize experience of attendees and develop sustainable service, this study applies IoT technology for stakeholders of convention.

In this paper, we seek to answer the following research questions: (1) What is the notion of sustainability and smartness in tourism field? (chapter 2-1) (2) What is the customer-driven interaction? (chapter 2-2) (3) What are the appropriate technologies to meet sustainable/smart service and the needs of smart convention and how to apply the technology? (chapter 2-2 and 3) (4) What do stakeholders such as companies and customers gain from sustainable and smart service for smart convention? (chapter 4)

To accomplish the objectives of the study, we took five phases of action research framework. In diagnosing phase, through interviews, observations, and literature studies, there searchers identify current issues and requirements for convention. In action planning phase, through the discussion with the practitioners of the convention, we have selected and arranged services for solving the problems. In action taking phase, smart buttons were installed in the COEX (Interaction convention place), we collected data during the convention period. In evaluation phase, we present the results and value of data. Finally, in specifying learning phase, the researchers summarized the benefits of each stakeholder.

The contribution of this paper is in four ways. First, the notion of sustainability was redefined by including not only socio-cultural, economy and environmental thinking but also the realms of customer-driven! interaction.

Second, this study proposed and constructed a new type of smart service with IoT for smart convention. Third, this research verified that benefits of sustainable and smart service. Finally, this can be used as an example of how conventions can find opportunities using new technologies.

An application of the loT technology based smart button was conducted for convention service, and the development and evaluation results are aligned lo action research framework including diagnosing, action planning, action taking, evaluation, and specifying learning phases. At the first and two phases, various difficulties and problems of the smart convention were diagnosed through interview with practitioners and service models were designed for solving the problems. Mobile service and appropriate IoT technology were discussed and applied to the convention space by installing smart button at the third phase. At the fourth, we derived in the role of Internet of things based mobile technology for smart convention and evaluated results of smart convention service in the various perspective. Finally, we proved value of mobile technology as the omni channel service through data analysis.

The Behavior Analysis of Exhibition Visitors using Data Mining Technique at the KIDS & EDU EXPO for Children

정민규, 김혜경, 최일영, 이경전, 김재경, 유아교육 박람회에서 데이터마이닝 기법을 이용한 전시 관람 행동 패턴 분석, 지능정보연구, 17(2), pp. 77-96, 2011. (PDF) in Korean

ABSTRACT

An exhibition is defined as market events for specific duration to present exhibitors’ main products to business or private visitors, and it plays a key role as effective marketing channels. As the importance of exhibition is getting more and more, domestic exhibition industry has achieved such a great quantitative growth. But, In contrast to the quantitative growth of domestic exhibition industry, the qualitative growth of Exhibition has not achieved competent growth. In order to improve the quality of exhibition, we need to understand the preference or behavior characteristics of visitors and to increase the level of visitors’ attention and satisfaction through the understanding of visitors. So, in this paper, we used the observation survey method which is a kind of field research to understand visitors and collect the real data for the analysis of behavior pattern. And this research proposed the following methodology framework consisting of three steps. First step is to select a suitable exhibition to apply for our method. Second step is to implement the observation survey method. And we collect the real data for further analysis. In this paper, we conducted the observation survey method to obtain the real data of the KIDS & EDU EXPO for Children in SETEC. Our methodology was conducted on 160 visitors and 78 booths from November 4th to 6th in 2010. And, the last step is to analyze the record data through observation. In this step, we analyze the feature of exhibition using Demographic Characteristics collected by observation survey method at first. And then we analyze the individual booth features by the records of visited booth. Through the analysis of individual booth features, we can figure out what kind of events attract the attention of visitors and what kind of marketing activities affect the behavior pattern of visitors. But, since previous research considered only individual features influenced by exhibition, the research about the correlation among features is not performed much. So, in this research, additional analysis is carried out to supplement the existing research with data mining techniques. And we analyze the relation among booths using data mining techniques to know behavior patterns of visitors. Among data mining techniques, we make use of two data mining techniques, such as clustering analysis and ARM(Association Rule Mining) analysis. In clustering analysis, we use K-means algorithm to figure out the correlation among booths. Through data mining techniques, we figure out that there are two important features to affect visitors’ behavior patterns in exhibition. One is the geographical features of booths. The other is the exhibit contents of booths. Those features are considered when the organizer of exhibition plans next exhibition. Therefore, the results of our analysis are expected to provide guideline to understanding visitors and some valuable insights for the exhibition from the earlier phases of exhibition planning. Also, this research would be a good way to increase the quality of visitor satisfaction. Visitors’ movement paths, booth location, and distances between each booth are considered to plan next exhibition in advance. This research was conducted at the KIDS & EDU EXPO for Children in SETEC(Seoul Trade Exhibition & Convention), but it has some constraints to be applied directly to other exhibitions. Also, the results were derived from a limited number of data samples. In order to obtain more accurate and reliable results, it is necessary to conduct more experiments based on larger data samples and exhibitions on a variety of genres.

Design of Serendipity Service Based on Near Field Communication Technology

이경전, 홍성우, NFC 기반 세렌디피티 시스템 설계, 지능정보연구, 17(4), 2011.(PDF) in Korean

ABSTRACT

The world of ubiquitous computing is one in which we will be surrounded by an ever-richer set of networked devices and services. Especially, mobile phone now becomes one of the key issues in ubiquitous computing environments. Mobile phones have been infecting our normal lives more thoroughly, and are the fastest technology in human history that has been adapted to people. In Korea, the number of mobile phones registered to the telecom company, is more than the population of the country. Last year, the numbers of mobile phone sold are many times more than the number of personal computer sold. The new advanced technology of mobile phone is now becoming the most concern on every field of technologies. The mix of wireless communication technology (wifi) and mobile phone (smart phone) has made a new world of ubiquitous computing and people can always access to the network anywhere, in high speed, and easily. In such a world, people cannot expect to have available to us specific applications that allow them to accomplish every conceivable combination of information that they might wish. They are willing to have information they want at easy way, and fast way, compared to the world we had before, where we had to have a desktop, cable connection, limited application, and limited speed to achieve what they want. Instead, now people can believe that many of their interactions will be through highly generic tools that allow end-user discovery, configuration, interconnection, and control of the devices around them. Serendipity is an application of the architecture that will help people to solve a concern of achieving their information. The word ‘serendipity’, introduced to scientific fields in eighteenth century, is the meaning of making new discoveries by accidents and sagacity. By combining to the field of ubiquitous computing and smart phone, it will change the way of achieving the information. Serendipity may enable professional practitioners to function more effectively in the unpredictable, dynamic environment that informs the reality of information seeking. This paper designs the Serendipity Service based on NFC (Near Field Communication) technology. When users of NFC smart phone get information and services by touching the NFC tags, serendipity service will be core services which will give an unexpected but valuable finding. This paper proposes the architecture, scenario and the interface of serendipity service using tag touch data, serendipity cases, serendipity rule base and user profile.

Bankruptcy Prediction Modeling Using Multiple Neural Network Models

Shin, K. Lee, K., Bankruptcy Prediction Modeling Using Multiple Neural Network Models, Lecture Notes in Artificial Intelligence 3214:668–674, September, 2004. – SCIE, ISSN:0302-9743

Abstract

The primary goal of this paper is to get over the limitations of single neural network models through model integration so as to increase the accuracy of bankruptcy prediction. We take the closeness of the output value to either 0 or 1 as the models confidence in its prediction as to whether or not a company is going to bankrupt. In case where multiple models yield conflicting prediction results, our integrated model takes the output value of the highest confidence as the final output. The output of the confidence-based integration approach significantly increases the prediction performance. The results of composite prediction suggest that the proposed approach will offer improved performance in business classification problems by integrating case-specific knowledge with the confidence information and general knowledge with the multi-layer perceptrons generalization capability.

Support Vector Machines Approach to Pattern Detection in Bankruptcy Prediction and its Contingency

Shin, K., Lee, K., Kim, H., Support Vector Machines Approach to Pattern Detection in Bankruptcy Prediction and its Contingency, Lecture Notes in Computer Science 3316:1254-1259, November, 2004.  – SCIE ISSN:0302-9743. pdf

Abstract

This study investigates the effectiveness of support vector machines (SVM) approach in detecting the underlying data pattern for the corporate failure prediction tasks. Back-propagation neural network (BPN) has some limitations in that it needs a modeling art to find an appropriate structure and optimal solution and also large training set enough to search the weights of the network. SVM extracts the optimal solution with the small training set by capturing geometric characteristics of feature space without deriving weights of networks from the training data. In this study, we show the advantage of SVM approach over BPN to the problem of corporate bankruptcy prediction. SVM shows the highest level of accuracies and better generalization performance than BPN especially when the training set size is smaller.