Visible to the public Sample-Based Regularization for Support Vector Machine Classification

TitleSample-Based Regularization for Support Vector Machine Classification
Publication TypeConference Paper
Year of Publication2017
AuthorsTran, D. T., Waris, M. A., Gabbouj, M., Iosifidis, A.
Conference Name2017 Seventh International Conference on Image Processing Theory, Tools and Applications (IPTA)
Keywordscomposability, Dropout, Fasteners, human action recognition tasks, image classification, Kernel, kernel methods, linear combination, maximum margin-based classification, Metrics, Optimization, pubcrawl, Regularization, regularization scheme, resilience, Resiliency, sample-based regularization, selected training data, support vector machine, support vector machine classification, Support vector machines, SVM, Training, Training data, training sample level, training sample selection

In this paper, we propose a new regularization scheme for the well-known Support Vector Machine (SVM) classifier that operates on the training sample level. The proposed approach is motivated by the fact that Maximum Margin-based classification defines decision functions as a linear combination of the selected training data and, thus, the variations on training sample selection directly affect generalization performance. We show that the exploitation of the proposed regularization scheme is well motivated and intuitive. Experimental results show that the proposed regularization scheme outperforms standard SVM in human action recognition tasks as well as classical recognition problems.

Citation Keytran_sample-based_2017