태양풍 매개변수를 이용한 Dst 예측 (Dst Prediction Based on Solar Wind Parameters)
-
- Journal of Astronomy and Space Sciences
- /
- 제26권4호
- /
- pp.425-438
- /
- 2009
ACE 위성에서 제공되는 1998년에서 2006년까지 9년간 관측된 양질의 1시간 평균 태양풍 매개변수를 이용하여 Burton et al.(1975)의 Dst 예측식을 재검토했다. 이 기간 동안 60개의 단순 발달형 자기폭풍을 골랐다. Burton et al.(1975)의 Dst 예측식의 에너지 공급항(Q)과 소멸 시간 (
Support vector machines(SVM)은 비교적 최근에 등장한 데이터마이닝 기법이지만, 재무, CRM 등의 경영학 분야에서 많이 연구되고 있다. SVM은 인공신경망과 필적할 만큼의 예측 정확도를 보이는 사례가 많았지만, 암상자로 불리는 인공신경망 모형에 비해 구축된 예측모형의 구조를 이해하기 쉽고, 인공신경망에 비해 과도적합의 가능성이 적어서 적은 수의 데이터에서도 적용 가능하다는 장점을 가지고 있다. 하지만, 일반적인 SVM을 이용하려면, 인공신경망과 마찬가지로 여러 가지 설계요소들을 설계자가 선택하여야 하기 때문에 임의성이 높고, 국부 최적해에 수렴할 가능성도 크다. 또한, 많은 수의 데이터가 존재하는 경우에는 데이터를 분석하고 이용하는데 시간이 소요되고, 종종 잡음이 심한 데이터가 포함된 경우에는 기대하는 수준의 예측성과를 얻지 못할 가능성이 있다. 본 연구에서는 일반적인 SVM의 장점을 그대로 유지하면서, 전술한 두 가지 단점을 보완한 새로운 SVM 모형을 제안한다. 본 연구에서 제안하는 모형은 사례선택기법을 일반적인 SVM에 융합한 것으로 대용량의 데이터에서 예측에 불필요한 데이터를 선별적으로 제거하여 예측의 정확도와 속도를 제고할 수 있는 방법이다. 본 연구에서는 잡음이 많고 예측이 어려운 것으로 알려진 재무 데이터를 활용하여 제안 모형의 유용성을 확인하였다.
Franchising is one of the fastest growing types of business. It is already popular and well-known in the U.S., and has been growing in many other countries including Korea. Furthermore, many Korean franchising companies have expanded their business overseas actively. According to the data by the Ministry of Industry and Resource, 82 companies out of a sample of 500 franchising companies are already operating in many foreign countries and 48% of them have started their foreign business since 2006. This clearly indicates the fast growing current trend of foreign operation by Korean franchising companies. In spite of the fast growing trend of foreign expansion in the industry, academic research on internationalization of franchising companies is extremely difficult to find. Accordingly, academic research on the issue is necessary and urgent in Korea. Among the various research questions on internationalization of franchising business, this study intends to investigate the difference in organizational factors between the franchising companies doing foreign operation and those doing business only domestically. More specifically, this research has the following purposes. First, considering the lack of theoretical basis of previous studies, resource-based theory and agency theory are employed as the theoretical bases. Second, this study explains the difference in internationalization based on organizational factors such as company size, history and growth rate. Third, the five hypotheses regarding the difference in organizational factors are presented and tested empirically, which is the first attempt in the area of this topic. Finally, the study attempts to clarify the conflicting implications among theories regarding some organizational factos such as growth rate. As the theoretical background, resource-based theory and agency theory are discussed. According to resource-based theory, a firm can grow continuously when it has competence and resource, and also the ability to develop them. The competence and resource can include capital, human resource, management skill, market information, ability to manage risk, etc. Meanwhile, agency theory views the relationship between franchisor and franchisee as an agency relationship. In agency theory, bonding capability and monitoring capability are the two key factors which promote internationalization of franchising companies. Based on the two theories, a conceptual model is designed. The model consists of two groups of variables. One is organizational factors including size, history, growth rate, price bonding and geographic dispersion. The other is whether a franchising company is operating overseas or not. We developed the following five research hypotheses basically describing the relationship between organizational factors and internationalization of franchising companies. H1: The size of franchising companies operating overseas is larger than that of franchising companies operating domestically. H2: The history of franchising companies operating overseas is longer than that of franchising companies operating domestically. H3: The growth rate of franchising companies operating overseas is higher than that of franchising companies operating domestically. H4: The price bonding of franchising companies operating overseas is higher than that of franchising companies operating domestically. H5: The geographic dispersion of franchising companies operating overseas is wider than that of franchising companies operating domestically. Data for the analyses are obtained from 2005 Korea Franchise Survey data co-generated by Ministry of Industry and Resource, GS1 Korea, and Korea Franchise Association. Out of 2,804 population companies, 2,489 companies are excluded for various reasons and 315 companies are selected as the final sample. Prior to hypotheses tests, validity and reliability of the measures of size, history, growth rate and price bonding are examined for further analyses. Geographic dispersion is not validated since it is measured using nominal data. A series of independent sample T-tests is used to find out whether there exists any significant difference between the companies internationalized and those operating only domestically for each organizational factor. Among the five factors, size and geographic dispersion show significant difference, growth rate and price bonding do not reveal any difference and, finally, history factor shows conflicting results in the difference depending on how to measure it.