Prediction of fatigue life of packaging EMC material based on RBF-SVM
by Hai Guo; Jinghua Yin; Jingying Zhao; Zhiyu Huang; Yue Pan
International Journal of Materials and Product Technology (IJMPT), Vol. 49, No. 1, 2014

Abstract: This paper applies a radial basis function support vector machine (RBF-SVM) for predicting fatigue life of packaging elastic memory composites (EMC) material. EMC is one of the three dominating materials in the FOL packing. Using EMC to encapsulate the large-scale integrated circuit has predominated research in the areas all over the world. Till now, more than 95% of microelectronic devices are PEMs. In this study, a RBF-SVM model was constructed to predict life of EMC material and the penalty parameter C and nuclear parameter R was optimised. Meanwhile, a comparison between the predictions of BP neural network, polynomial kernel, sigmoid kernel, and RBF kernel was made. Experiments show that the SVM (RBF-SVM) has higher complexity and prediction accuracy than polynomial kernel and sigmoid kernel have, and its prediction accuracy is far higher than that of BP neural network model. According to the predictions, RBF-SVM is very suitable for applying in the prediction of fatigue life of packaging EMC material and receives good prediction complexity and accuracy.

Online publication date: Sat, 19-Jul-2014

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Materials and Product Technology (IJMPT):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com