Using salient features to reinforce GA-SVM for business crisis diagnoses

Liang-Hsuan Chen, Huey Der Hsiao

Research output: Contribution to journalArticlepeer-review

6 Citations (Scopus)


This research considers the benefits of integrating a real-valued genetic algorithm and saliency analysis into an SVM for the determination of parameter values and the selection of salient features to carry out the diagnosis of a business crisis. Two phases, corresponding to the financial and intellectual capital indices used in real businesses, are adopted to determine the salient features. Eight salient features, including six financial indices and two intellectual capital ones, are used for the diagnosis. These features are common and easily accessible from publicly available information, making the proposed method very practical for business managers who wish to conduct a real-time investigation of the potential for a business crisis. A series of learning and testing processes with real business data show that the diagnosis model has a crisis prediction accuracy of up to 96.11%, demonstrating the applicability of the proposed method.

Original languageEnglish
Pages (from-to)4487-4501
Number of pages15
JournalInternational Journal of Innovative Computing, Information and Control
Issue number10
Publication statusPublished - 2010 Oct

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Software
  • Information Systems
  • Computational Theory and Mathematics

Fingerprint Dive into the research topics of 'Using salient features to reinforce GA-SVM for business crisis diagnoses'. Together they form a unique fingerprint.

Cite this