If-SVM: Iterative factoring support vector machine
- PDF / 2,068,822 Bytes
- 21 Pages / 439.642 x 666.49 pts Page_size
- 45 Downloads / 218 Views
If-SVM: Iterative factoring support vector machine Yuqing Pan1 · Wenpeng Zhai1 · Wei Gao1 · Xiangjun Shen1 Received: 24 July 2019 / Revised: 31 May 2020 / Accepted: 4 June 2020 / © Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract Support Vector Machine (SVM) is widely applied in classification and regression tasks where support vectors are pursued through convex quadratic programming technique due to its effectiveness and efficiency. However, existing studies ignore the importance of training samples when they are fed into the model. In this paper, we propose a novel Iterative Factoring Support Vector Machine (If-SVM) method. Sample factoring is introduced in our proposed model to measure the significance of each data point, where it can effectively reduce the negative impact of trivial or noisy data points. In this way, our proposed model is concentrates on the critical data points falling around the hyperplane. By introducing this iterative factoring of data points into SVM, the classification accuracy of our proposed method is above that of 1.45% than other comparative methods in image recognition datasets. Experimental results on a variety of UCI demonstrate that, our proposed method has superior performances in decreasing the total number of support vectors than the other state-of-the-art SVM methods. More importantly, our further experiments also illustrate that, the classification performance of the state-of-the-art SVM methods can be improved 1.29% by incorporating our sample factoring idea into their models, which demonstrate our idea is a useful tool to improve the state-of-art SVM models. Keywords Support vector machine (SVM) · Factoring · Iterative · Noisy samples
1 Introduction Support Vector Machine (SVM) based on statistical learning theory is a machine learning algorithm proposed by Vapnik in [3, 4, 10]. The SVM seeks an optimal separation hyperplane between limited positive and negative sample information, and to find the optimal compromise between the complexity of the model and generalization ability has shown its advantages of effectiveness and efficiency in classification and regression task with support vectors being pursued through convex quadratic programming technique[2, 20, 31].
Xiang-Jun Shen
[email protected] 1
School of Computer Science and Communication Engineering, JiangSu University, JiangSu, 212013, China
Multimedia Tools and Applications
In order to improve the classification performance of SVM, various improved methods were proposed subsequently. One of such the methods applied mutual information (MI) to measure the relevance between two random variables [18, 24, 25], and to estimate the MI between each feature and the given class labels [12, 22]. The weights of each feature estimated by the MI method improve the generalization ability of the traditional SVMs, whereas show bad performance in high dimensions. Therefore, a novel radius-margin-based SVM model for joint learning of feature transformation and the SVM classifier [7, 21, 26– 29] was p
Data Loading...
