2009 | OriginalPaper | Chapter
SVM Based Decision Analysis and Its Granular-Based Solving
Authors : Tian Yang, Xinjie Lu, Zaifei Liao, Wei Liu, Hongan Wang
Published in: Computational Science and Its Applications – ICCSA 2009
Publisher: Springer Berlin Heidelberg
Activate our intelligent search to find suitable subject content or patents.
Select sections of text to find matching patents with Artificial Intelligence. powered by
Select sections of text to find additional relevant content using AI-assisted search. powered by
This paper proposes an approach to decision analysis for complex industrial process without enough knowledge of input-output model, which is based on the two-class SVM method. It first proposes a SVM Based Decision Analysis Model to improve the accuracy of determinant of whether a decision is acceptable/ unacceptable by verifying the ‘soft margin’ of a SVM. This makes it only allow misclassification of only one class in a two-class classification. Then a granular-based approach is presented to solving this model. It is proved that this granular approach can reach an upper bound of the original SVM model. An algorithm then is presented to determine whether a decision is acceptable. According to our analysis and experiments, the two types of SVM have better accuracy on judging its target class then traditional SVM, and the granular-based SVM solving can reduce the running time.