Abstract
In consideration of the problem that different features have different contributions to recognition results in pattern recognition for composite plates bonding flaws, this paper proposes feature weighting SVM algorithm based on information gain. The algorithm firstly calculates the contributions of all features to recognition results by means of information gain, then weights kernel function in recognition and decision making via the results combining with Support Vector Machines in order to emphasize the roles of important features in the process of recognition and eliminate or alleviate the influence of wake-relevant and irrelevant features on recognition results. Both theory analysis and experience show that this algorithm can be used to recognize bonding flaw of composite plate accurately and quantitatively.
Keywords
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
Zhao, Y.-G., Zhang, Z.: Study of ultrasonic detection for bonding flaws of thin plate based on time-frequency analysis and artificial neural network, pp. 27–33. Inner Mongolia University (2008)
Xu, Y.-H., Zhang, Z.: Study on the fuzzy neural network pattern recognition method for ultrasonic detection on bonding detect of thin composite materials, pp. 26–36. Inner Mongolia University (2009)
Wang, Z.-J., Zhou, R.-J.: Application of improved FCM clustering algorithm in felting quality ultrasonic detection of composite material. In: Proceedings of the 29th Chinese Control Conference, pp. 2953–2956 (2010)
Zhu, L.-Y., Cao, C.-X.: The recognition of flaws based on Support Vector Machine. Journal of ChongQing University 25(6), 42–45 (2002)
Du, S.-X., Wu, T.-J.: Support vector machines for pattern recognition. Journal of ZheJiang University 37(5), 521–526 (2003)
Zhai, Y.-J.: Research on fault intelligent diagnosis based on Support Vector Machines, pp. 42–57. North China Electric Power University (2004)
Zhang, X.-G.: Introduction to statistical learning theory and Support Vector Machines. Automatica Sinica 26(1), 33–41 (2000)
Chen, Z.-Z., Li, L., Yao, Z.-G.: Feature weighting KNN based on SVM. Scientiarum Naturalium Universitatis Sunyatseni 44(1), 17–20 (2005)
Lin, Y.-M., Lu, Z.-Y., Zhao, S., Zhu, W.-D.: The study of feature weighting for vector space model. Journal of Information 3, 5–7 (2008)
Wang, T.-H., Tian, S.-F., Huang, H.-K.: Feature Weighted Support Vector Machine. Journal of Electronics & Information Technology 31(3), 514–517 (2009)
Wang, H.-T., Zhang, Z.: Research of feature extraction for flaw ultrasonic echo based on wavelet Pack analysis. ICMLC, V2-68–V2-72 (2011)
Trafalis, T.B., Oladunni, O., Papavassiliou, D.V.: Two-Phase Flow Regime Identification with a Multi-classification on Support Vector Machine (SVM) Model. Ind. Eng. Chem. Res. 44, 4414–4426 (2005)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2012 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Zhang, Z., Wang, Xf., Wang, Ht., Liu, Yx. (2012). Research on Quantitative Recognition for Composite Plate Bonding Flaw Based on Feature Weighting SVM. In: Zhao, M., Sha, J. (eds) Communications and Information Processing. Communications in Computer and Information Science, vol 288. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-31965-5_42
Download citation
DOI: https://doi.org/10.1007/978-3-642-31965-5_42
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-31964-8
Online ISBN: 978-3-642-31965-5
eBook Packages: Computer ScienceComputer Science (R0)