다수의 인공신경망 모형을 통합한 기업부도 예측모형에 관한 연구

신경식(Kyung Shik Shin) 이화여자대학교 경영연구소 2000 신산업경영저널 Vol.18 No.1

데이터 마이닝 기법을 활용한 스팸 메일 분류 및 예측모형 구축에 관한 연구

신경식(Kyung Shik Shin),안수산(Su San Ahn) 이화여자대학교 경영연구소 2002 신산업경영저널 Vol.20 No.2

美國 人身拘束節次上 適正한 訴追權 行使와 人權保護의 調和

申勁植(Kyung-Shik Shin) 한국법학원 1999 저스티스 Vol.32 No.1

An Integrated Approach Using CBR and Induction Technique for Bond Rating Models

신경식(Kyung Shik Shin) 이화여자대학교 경영연구소 2000 신산업경영저널 Vol.18 No.2

가상 팀의 교류활성기억 시스템과 팀 성과의 관계

신경식(Kyung-shik Shin),서아영(Ayoung Suh) 한국전자거래학회 2010 한국전자거래학회지 Vol.15 No.2

가상 팀은 구성원들이 지리적으로 떨어져 있고 기술로 매개된 환경하에서 주로 협업하기 때문에 대면접촉을 위주로 하는 전통적인 팀에 비하여 팀의 교류활성기억 시스템 (Transactive Memory System:TMS) 구축이 더욱 중요한 것으로 간주되고 있다. 본 연구의 목적은 TMS 이론을 중심으로 이론적 고찰 및 선행 변수를 탐색함으로써, 가상 팀의 효과적인 운영을 위한 학문적 실무적 시사점을 제공하는 것이다. 이를 위하여 문헌조사를 통하여 가상 팀의 성과를 예측하는 이론적 모형을 개발하였다. 먼저 TMS를 구성하는 세가지 요인, 즉 (1) 전문성 파악, (2) 업무조정, (3) 인지기반 신뢰를 도출하였다. 그리고 TMS의 선행변인으로는 (1) 지각된 매체 풍부성, (2) 네트워크 연결강도, (3) 공유된 규범, (4) 지리적 거리등을 도출하였다. 제시된 이론적 모형을 검증하기 위하여 179개의 가상 팀을 대상으로 경로분석을 실시하였다. 분석결과, 가상 팀의 TMS를 구성하는 세가지 요인, 전문성 파악, 업무조정, 인지기반 신뢰는 가상 팀의 성과를 예측하는 결정적인 요인인 것으로 나타났다. 또한 TMS의 선행변인으로 제시된 지각된 매체 풍부성은 인지기반 신뢰에 유의한 정의 영향을 미치며, 네트워크 연결강도는 업무조정과 인지기반 신뢰에 유의한 정의 영향을 미치는 것으로 나타났다. 반면, 공유된 규범은 가상 팀의 교류활성기억시스템 구성 요인 세 가지 모두에 유의한 정의 영향을 미치는 것으로 나타났다. 예상과 달리, 구성원들 간의 지리적 거리는 TMS 세 가지 구성 요소 모두에 유의한 영향을 미치지 않았다. A virtual team is defined a group of people that use electronic communications for some or all of their interactions with other team members. Because team members of a virtual team are physically and temporally distributed, a team’s transactive memory system (TMS) is considered to be crucial for the team’s effectiveness and performance. TMS refers to a set of individual memory systems which integrate knowledge possessed by particular members through a shared awareness of who knows what. This paper seeks to understand (1) how a virtual team’s TMS influences team performance, and (2) what factors contribute to developing the team’s TMS. Given these purposes, through the extensive literature review, we first identified components and antecedents to develop a theoretical model that predicts a virtual team’s performance. Using the survey data gathered from 172 virtual teams, this study found that expertise location, coordination, and cognition-based trust which were proposed as three components of TMS positively influenced a virtual team’s performance. Furthermore, this study uncovered that perceived media richness, network tie strength, and shared norms significantly influenced the components of TMS, while geographical dispersion did not exert any significant influence on TMS.

A Neuro-genetic Approach for Bankruptcy Prediction Modeling

신경식(Kyung Shik Shin) 이화여자대학교 경영연구소 1999 신산업경영저널 Vol.17 No.2

유비쿼터스 환경에서 개인자산관리 서비스를 위한 지능형 에이전트의 설계

신경식(Kyung-shik Shin),김남희(Nam-hee Kim) 한국지능정보시스템학회 2009 지능정보연구 Vol.15 No.4

The rapid changes of financial environment have increased the need and demand for personal financial advisory service from financial experts. In particular, as the individual customers want to get more customized financial services, the financial institutions created the private banking (PB) sector and have constantly expanded their PB services. However, it remains still problematic that the private banking system requires high costs so that the number of eligible customers who can have proper PB services is quite limited. To solve this problem, we propose an intelligent agent that can provides specialized and customized personal financial advisory services to the customers with low costs. The proposed agent systemizes and structures the information and knowledge of financial experts in private banking services so that individual customers can easily access to high?quality PB services when they need. On the first attempt we develop a framework of U?smart PB, an intelligent agent for personal financial management based on different scenarios related to personal financial decisions, and derive its core services. This system not only provides information simply, but also proposes to support personal investment decisions technically as an intelligent agent that embodies real?time customized financial management in a ubiquitous environment, regardless of time and place.

Development of the Knowledge-based Systems for Anti-money Laundering in the Korea Financial Intelligence Unit

Kyung-shik Shin(신경식),Hyun-jung Kim(김현정),Hyo-sin Kim(김효신) 한국지능정보시스템학회 2008 지능정보연구 Vol.14 No.2

This case study shows constructing the knowledge-based system using a rule-based approach for detecting illegal transactions regarding money laundering in the Korea Financial Intelligence Unit (KoFIU). To better manage the explosive increment of low risk suspicious transactions reporting from financial institutions, the adoption of a knowledge-based system in the KoFIU is essential. Also since different types of information from various organizations are converged into the KoFIU, constructing a knowledge-based system for practical use and data management regarding money laundering is definitely required. The success of the financial information system largely depends on how well we can build the knowledge-base for the context. Therefore we designed and constructed the knowledge-based system for anti-money laundering by committing domain experts of each specific financial industry co-worked with a knowledge engineer. The outcome of the knowledge base implementation, measured by the empirical ratio of Suspicious Transaction Reports (STRs) reported to law enforcements, shows that the knowledge-based system is filtering STRs in the primary analysis step efficiently, and so has made great contribution to improve efficiency and effectiveness of the analysis process. It can be said that establishing the foundation of the knowledge base under the entire framework of the knowledge-based system for consideration of knowledge creation and management is indeed valuable.

가상팀의 지식네트워크, IT 활용, 성과 간의 관계에 관한 연구

신경식(Kyung Shik Shin),서아영(A Young Suh) 한국경영학회 2009 經營學硏究 Vol.38 No.1

A virtual team is defined as a group of people that works across space, time, and organizational boundaries with links strengthened by the diverse computer-mediated communication media. By forming virtual teams, people can easily collaborate with other people and acquire information and knowledge from anywhere and at anytime. However, managing virtual teams is far more challenging than managing traditional teams because ad hoc virtual teams may engage in a lower level of social interacts and supports than their counterparts in face-to-face environments. The problem of how to manage social relationships among virtual team members to increase their performance is considered to be very pivotal and still remains unsolved. Through the lens of social networks, this study examines the following specific questions: (1) to what extent should ad hoc virtual teams maintain internal cohesion; (2) whether do virtual team members’ external bridging ties guarantee the higher level of team performance; and (3) how does IT play the role in balancing and harmonizing internal cohesion and external bridging ties for increasing the performance of a virtual team. We investigate these questions with the specific type of social networks; the knowledge networks which contain knowledge or task-related intelligence in the networks. These knowledge networks contain the collective competencies that enable organizational members to produce products and services. By examining the influence of knowledge networks of virtual teams on their performance, this study aims to reveal how to manage knowledge network among members to maximize the performance of a virtual team. The contributions of this study are three-fold. First, we conceptually clarify the properties of knowledge networks that individuals maintain within and outside their work groups: network density and group-level structural holes. Next, by integrating social network theory and CMC related theories, we develop a theoretical framework linking knowledge networks, IT use, and team performance. Third, we examine the interplay among the research constructs and elucidate a complex phenomenon in relations to the forming and managing knowledge networks of a virtual team, which is in short supply in the current literature. The proposed model was tested with 172 consultants and 42 virtual teams in 5 global consulting companies in Korea. The study found that increasing network intra-group network density was crucial for better team performance. On the contrary, increasing extra-group structural holes decreased the team performance, contradicting the prevailing wisdom. The most important standpoint is that the effect of knowledge network of virtual teams varied along the level of IT use that is related to communal CMC such as group ware or group decision support systems (GDSS). This study is meaningful in that it examined how to manage knowledge network by integrating social network theory and computer-mediation communication theory, and by analyzing the moderating effect between knowledge network and IT use in the context of virtual teams. Based on these findings, the study suggests that optimal knowledge network configurations to maximize the performance of a virtual team should be considered along with the IT use.

Business Application of Convolutional Neural Networks for Apparel Classification Using Runway Image

Yian Seo(서이안),Kyung-shik Shin(신경식) 한국지능정보시스템학회 2018 지능정보연구 Vol.24 No.3

Large amount of data is now available for research and business sectors to extract knowledge from it. This data can be in the form of unstructured data such as audio, text, and image data and can be analyzed by deep learning methodology. Deep learning is now widely used for various estimation, classification, and prediction problems. Especially, fashion business adopts deep learning techniques for apparel recognition, apparel search and retrieval engine, and automatic product recommendation. The core model of these applications is the image classification using Convolutional Neural Networks (CNN). CNN is made up of neurons which learn parameters such as weights while inputs come through and reach outputs. CNN has layer structure which is best suited for image classification as it is comprised of convolutional layer for generating feature maps, pooling layer for reducing the dimensionality of feature maps, and fully-connected layer for classifying the extracted features. However, most of the classification models have been trained using online product image, which is taken under controlled situation such as apparel image itself or professional model wearing apparel. This image may not be an effective way to train the classification model considering the situation when one might want to classify street fashion image or walking image, which is taken in uncontrolled situation and involves people’s movement and unexpected pose. Therefore, we propose to train the model with runway apparel image dataset which captures mobility. This will allow the classification model to be trained with far more variable data and enhance the adaptation with diverse query image. To achieve both convergence and generalization of the model, we apply Transfer Learning on our training network. As Transfer Learning in CNN is composed of pre-training and fine-tuning stages, we divide the training step into two. First, we pre-train our architecture with large-scale dataset, ImageNet dataset, which consists of 1.2 million images with 1000 categories including animals, plants, activities, materials, instrumentations, scenes, and foods. We use GoogLeNet for our main architecture as it has achieved great accuracy with efficiency in ImageNet Large Scale Visual Recognition Challenge (ILSVRC). Second, we fine-tune the network with our own runway image dataset. For the runway image dataset, we could not find any previously and publicly made dataset, so we collect the dataset from Google Image Search attaining 2426 images of 32 major fashion brands including Anna Molinari, Balenciaga, Balmain, Brioni, Burberry, Celine, Chanel, Chloe, Christian Dior, Cividini, Dolce and Gabbana, Emilio Pucci, Ermenegildo, Fendi, Giuliana Teso, Gucci, Issey Miyake, Kenzo, Leonard, Louis Vuitton, Marc Jacobs, Marni, Max Mara, Missoni, Moschino, Ralph Lauren, Roberto Cavalli, Sonia Rykiel, Stella McCartney, Valentino, Versace, and Yve Saint Laurent. We perform 10-folded experiments to consider the random generation of training data, and our proposed model has achieved accuracy of 67.2% on final test. Our research suggests several advantages over previous related studies as to our best knowledge, there haven’t been any previous studies which trained the network for apparel image classification based on runway image dataset. We suggest the idea of training model with image capturing all the possible postures, which is denoted as mobility, by using our own runway apparel image dataset. Moreover, by applying Transfer Learning and using checkpoint and parameters provided by Tensorflow Slim, we could save time spent on training the classification model as taking 6 minutes per experiment to train the classifier. This model can be used in many business applications where the query image can be runway image, product image, or street fashion image. To be specific, runway query image can be used for mobile application service during fashion week to facilitate brand search, street style query image