[1]高雅田,贾斯淇.基于Gram-Schmidt正交化和HSIC的核函数选择方法[J].计算机技术与发展,2024,34(06):148-154.[doi:10.20165/j.cnki.ISSN1673-629X.2024.0086]
　GAO Ya-tian,JIA Si-qi.A Kernel Selection Method Based on Gram-Schmidt Orthogonalization and HSIC[J].,2024,34(06):148-154.[doi:10.20165/j.cnki.ISSN1673-629X.2024.0086]
点击复制

基于Gram-Schmidt正交化和HSIC的核函数选择方法()

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
34
期数:
2024年06期
页码:
148-154
栏目:
人工智能
出版日期:
2024-06-10

文章信息/Info

Title:
A Kernel Selection Method Based on Gram-Schmidt Orthogonalization and HSIC
文章编号:
1673-629X(2024)06-0148-07
作者:
高雅田; 贾斯淇
东北石油大学 计算机与信息技术学院,黑龙江 大庆 163318
Author(s):
GAO Ya-tian; JIA Si-qi
Dept. of Computer and Information Technology,Northeast Petroleum University,Daqing 163318,China
关键词:
多核学习; 核函数选择; 不相关冗余信息; Gram-Schmidt正交化; Hilbert-Schmidt独立准则
Keywords:
multiple kernel learning;  kernel selection;  irrelevant redundant information;  Gram - Schmidt orthogonalization;  Hilbert - Schmidt independence criterion
分类号:
TP301
DOI:
10.20165/j.cnki.ISSN1673-629X.2024.0086
摘要:
核方法是一种解决非线性、异构数据的有效方法,核函数的选择问题是核方法中的一个重要课题,对于不同的应用问题,如何选择合适的核函数还没有足够的理论基础,不适当的核函数选取会降低核方法的性能。 由此,提出了一种基于 Gram-Schmidt 正交化(GSO)和 Hilbert-Schmidt 独立准则的核选择方法(HSIC-GSO),该方法考虑了核函数选择过程中存在的不相关冗余信息。 首先,利用 GSO 消除核函数之间的冗余信息;然后,使用 HSIC 度量核函数与理想核之间的相似性;最后,得到一组判别能力强、多样性大的基核函数。 实验结果表明,HSIC-GSO 方法选择的核函数泛化性好,并且提高了 MKL 的分类性能,验证了所提方法的有效性。
Abstract:
Kernel methods are an effective method for solving nonlinear and heterogeneous data,and the selection of kernel functions is an important issue in kernel methods. For different application problems,there is not enough theoretical basis for selecting the appropriate kernel,and improper kernel selection can degrade the performance of kernel methods. A kernel selection method based on Gram-Schmidt orthogonalization ( GSO) and Hilbert - Schmidt independence criterion ( HSIC - GSO) is proposed, which considers the irrelevant redundant information present in the kernel function selection process. Firstly,GSO is used to eliminate redundant information between kernel functions. Then,HSIC is used to measure the similarity between the kernel function and the ideal kernel. Finally,a set of kernel with strong discriminative ability and high diversity is obtained. The experimental results show that the HSIC-GSO method has good kernel generalization and improves the classification performance of MKL,verifying the effectiveness of the proposed method.

相似文献/References:

[1]江伟,潘昊.基于优化的多核学习方法的Web文本分类的研究[J].计算机技术与发展,2013,(10):80.
　JIANG Wei[],PAN Hao[].Research of Web Document Classification Based on Optimized Multiple Kernel Learning Method[J].,2013,(06):80.
[2]黄 琳,荆晓远,董西伟.基于多核集成学习的跨项目软件缺陷预测[J].计算机技术与发展,2019,29(06):27.[doi:10. 3969 / j. issn. 1673-629X. 2019. 06. 006]
　HUANG Lin,JING Xiao-yuan,DONG Xi-wei.Cross-project Software Defect Prediction Based on Multiple Kernel Ensemble Learning[J].,2019,29(06):27.[doi:10. 3969 / j. issn. 1673-629X. 2019. 06. 006]
[3]王梅,任怡果*,刘勇,等.求解多核学习的自适应随机递归梯度下降法[J].计算机技术与发展,2025,(07):93.[doi:10.20165/j.cnki.ISSN1673-629X.2025.0038]
　WANG Mei,REN Yi-guo*,LIU Yong,et al.Adaptive Stochastic Recursive Gradient Algorithm for Multiple Kernel Learning[J].,2025,(06):93.[doi:10.20165/j.cnki.ISSN1673-629X.2025.0038]

更新日期/Last Update: 2024-06-10