International Journal of
Physical Sciences

  • Abbreviation: Int. J. Phys. Sci.
  • Language: English
  • ISSN: 1992-1950
  • DOI: 10.5897/IJPS
  • Start Year: 2006
  • Published Articles: 2568

Full Length Research Paper

Forecasting the output of Taiwan’s integrated circuit (IC) industry using empirical mode decomposition and support vector machines

Kwo-Liang Chen1, Ching-Chiang Yeh2* and Tz-ling Lu3
  1Department of Industrial Engineering and Management, China University of Science and Technology, No.245, Sec. 3, Academia Rd., Nangang Dist., Taipei City, Taiwan 115, R.O.C. 2Department of Business Administration, National Taipei College of Business, No.321, Sec. 1, Ji-Nan Rd., Zhongzheng District, Taipei City, Taiwan 10051, R.O.C. 3Department of Business Administration, Soochow University, No.56. Kueiyang Street, Section 1, Taipei, Taiwan 100, R.O.C.
Email: [email protected]

  •  Accepted: 10 August 2012
  •  Published: 09 October 2012

Abstract

 

As the production values in the integrated circuit (IC) industry are inherently nonlinear and non-stationary, it is regarded as one of the most challenging tasks for practitioners and academics. This study proposed a hybrid methodology by combining empirical mode decomposition (EMD) and support vector regression (SVR) in production values forecasting. The proposed approach first uses EMD, which can adaptively decompose the complicated raw data into a finite set of intrinsic mode functions (IMFs) and a residue. After identifying the IMF components, residue are then modeled and forecasted using SVR. Thefinal forecasting value can be obtained by the sum of these prediction results. Experimental results show that the proposed approach outperforms the SVR model without EMD preprocessing.

 

Key words: Integrated circuit (IC) industry, production values forecasting, empirical mode decomposition, support vector regression.