E3 Journal of Business Management and Economics
E3 Journal of Business Management and Economics Vol. 3 (2) pp. 079-089, February 2012; © E3 Journals; ISSN 2141-7482
Construction of the prediction model of business operation performance in the electronic industry
Shing-Chih Yang * , Chan-Shal Lee , Hsuan-Shih LeeShipping and Transportation Management Department, National Taiwan Ocean University, Taiwan, R.O.C. No.2, 2 Pei-Ning Road,Keelung ,Taiwan 20224,R.O.C.
*Corresponding Author E-mail: scyang2012@gmail.com
Accepted 10 January 2012
Abstract
No more than three years after the occurrence of sub-prime crisis in USA, the global financial market is facing again the strict threat of the European Debt Crisis, and if it cannot be solved, its impact will be far beyond financial crisis. Fast change of economic cycle has led to financial difficulty and credit bankruptcy in lots of enterprises in Taiwan with bad business operation performance; hence, lots of enterprises have to implement no-pay day or even lay-off. In addition, the confidence of lots of stock investors is affected, and many of them hesitate to buy more stocks. First, this article collected two sets of financial report data from 350 electronic related companies with stocks listed in regular and OTC stock market in Taiwan in the second season of 2011 and third season of 2011, meanwhile, Grey Relational Analysis and Data Envelopment Analysis where used to investigate the business operation performance of each enterprise and to rank the result. Then, dichotomy method was used in this article to divide this ranking into two types of good and bad performance to be used as dependent variable (Y), meanwhile, the financial report data was collected and arranged to be used as independent variable (X), then models such as ZSCORE model (abbreviated as ZSCORE), the association of Support Vector Regression and ZSCORE model (abbreviated as SVR+ZSCORE), Artificial Fish Swarm Algorithm (AFSA) optimized Support Vector Regression parameter in association with ZSCORE model (abbreviated as AFSASVR+ZSCORE) and Fruit Fly Optimization Algorithm optimized Support Vector Regression parameter in association with ZSCORE model (abbreviated as FOASVR+ZSCORE), will be used respectively to set up enterprise’s operation performance detection model, meanwhile, the classification prediction capability of each model will be compared, and the result will be provided to the business operation and management personnel as reference.
Keywords: Grey Relational Analysis; ZSCORE; Artificial Fish Swarm Algorithm; Fruit Fly Optimization Algorithm; Support Vector Regression
[Download Article - PDF]