基于多核学习支持向量机的音乐流派分类

doi:10.11772/j.issn.1001-9081.2015.06.1753

计算机应用 ›› 2015, Vol. 35 ›› Issue (6): 1753-1756.DOI: 10.11772/j.issn.1001-9081.2015.06.1753

基于多核学习支持向量机的音乐流派分类

孙辉^1,2, 许洁萍^1,2, 刘彬彬¹

1. 中国人民大学信息学院, 北京 100872;
2. 数据工程与知识工程教育部重点实验室(中国人民大学), 北京 100872

收稿日期:2015-01-13 修回日期:2015-04-17 发布日期:2015-06-12
通讯作者: 许洁萍(1966-),女,黑龙江牡丹江人,副教授,博士,CCF会员,主要研究方向:多媒体计算、音频信息处理;xjieping@ruc.edu.cn
作者简介:孙辉(1977-),女,山东平度人,讲师,博士,CCF会员,主要研究方向:音频信息处理、数据库;刘彬彬(1990-),男,山东枣庄人,硕士,主要研究方向:音乐信息分类。
基金资助:
中国人民大学科学研究基金(中央高校基本科研业务费专项资金)资助项目(14XNLQ01)。

Music genre classification based on multiple kernel learning and support vector machine

SUN Hui^1,2, XU Jieping^1,2, LIU Binbin¹

1. School of Information, Renmin University of China, Beijing 100872, China;
2. Key Laboratory of Data Engineering and Knowledge Engineering, Ministry of Education (Renmin University of China), Beijing 100872, China

Received:2015-01-13 Revised:2015-04-17 Online:2015-06-12

摘要/Abstract

摘要：

针对不同特征向量下选择最优核函数的学习方法问题,将多核学习支持向量机(MK-SVM)应用于音乐流派自动分类中,提出了将最优核函数进行加权组合构成合成核函数进行流派分类的方法。多核分类学习能够针对不同的声学特征采用不同的最优核函数,并通过学习得到各个核函数在分类中的权重,从而明确各声学特征在流派分类中的权重,为音乐流派分类中特征向量的分析和选择提供了一个清晰、明确的结果。在ISMIR 2011竞赛数据集上验证了提出的基于多核学习支持向量机(MKL-SVM)的分类方法,并与传统的基于单核支持向量机的方法进行了比较分析。实验结果表明基于MKL-SVM的音乐流派自动分类准确率比传统单核支持向量机的分类准确率提高了6.58%,且该方法与传统的特征选择结果比较,更清楚地解释了所选择的特征向量对流派分类的影响大小,通过选择影响较大的特征组合进行分类,分类结果也有了明显的提升。

关键词: 音乐流派分类, 多核学习, 支持向量机, 特征选择, 模式识别

Abstract:

Multiple Kernel Learning and Support Vector Machine (MKL-SVM) was applied to automatic music genre classification to choose the optimal kernel functions for different features, a method of conducting the optimal kernel function combination into the synthetic kernel function by weighting for music genre classification was proposed. Different optimal kernel functions were chosen for different acoustic features by multiple kernel classification learning, the weight of each kernel function in classification was obtained, and the weight of each acoustic feature in the classification of the genre was clarified, which provided a clear and definite result for the analysis and selection of the feature vector in the classification of music genre. The experiments on the dataset of ISMIR 2011 show that, compared with the traditional single kernel support vector machine classification, the accuracy of the proposed music genre automatic classification method based on MKL-SVM is greatly improved by 6.58%. And the proposed method can more clearly reveal the the different features' impacts on music genre classification results, the classification results has also been significantly improved by selecting features with larger effects on classification.

Key words: music genre classification, multiple kernel learning, Support Vector Machine (SVM), feature selection, pattern recognition

中图分类号:

TP391.4

孙辉, 许洁萍, 刘彬彬. 基于多核学习支持向量机的音乐流派分类[J]. 计算机应用, 2015, 35(6): 1753-1756.

SUN Hui, XU Jieping, LIU Binbin. Music genre classification based on multiple kernel learning and support vector machine[J]. Journal of Computer Applications, 2015, 35(6): 1753-1756.

参考文献

[1] TZANETAKIS G, COOK P. Musical genre classification of audio signals [J]. IEEE Transactions on Speech and Audio Processing, 2002, 10(5): 293-302.
[2] BACH F R, LANCKRIET G R G, JORDAN M I. Multiple kernel learning, conic duality, and the SMO algorithm [EB/OL].[2014-12-02]. http://www.cs.berkeley.edu/~jordan/papers/skm_icml.pdf.
[3] LUKASHEVICH H. Applying multiple kernel learning to automatic genre classification [C]//Proceedings of the 34th Annual Conference of Challenges at the Interface of Data Analysis, Computer Science, and Optimization.Berlin: Springer, 2012: 393-400.
[4] VAPNIK V N. An overview of statistical learning theory [J]. IEEE Transactions on Neural Network, 1999, 10(5): 988-999.
[5] SONNENBURG S, RÄTSCH G, SCHÄFER C, et al. Large scale multiple kernel learning [J]. Journal of Machine Learning Research, 2006, 7: 1531-1565.
[6] KOSTEK B, KUPRYJANOW A, ZWAN P, et al. Report of the ISMIS 2011 contest: music information retrieval [C]//ISMIS 2011: Proceedings of the 19th International Symposium on Foundations of Intelligent Systems, LNCS 6804. Berlin: Springer, 2011: 715-724.
[7] DING Y, YAN Z, GAO Z. A method for selecting parameter of SVM with RBF kernel [EB/OL]. Beijing:Chinese Sciencepaper Online. [2014-12-05]. http://www.paper.edu.cn/html/releasepaper/2010/05/659/. (丁允静, 闫志刚, 高瞻.一种RBF核SVM的参数选择方法[EB/OL].北京:中图科技论文在线. [2014-12-05].http://www.paper.edu.cn/html/releasepaper/2010/05/659/.)
[8] ZENG C, SONG S, XU J. Study on multi-modal music genre classification [J]. Journal of Frontiers of Computer Science and Technology, 2011, 5(1): 50-58 (甄超, 宋爽, 许洁萍.多模态音乐流派分类研究[J].计算机科学与探索, 2011, 5(1):50-58.)
[9] WVLFING J. RIDEMILLER M. Unsupervised learning of local features for music classification [EB/OL]. [2014-12-03]. http://ml.informatik.uni-freiburg.de/_media/publications/wuelf2012.pdf.
[10] XU J. WANG P. YAN L. Feature selection for automatic classification of Chinese folk songs [C]//CISP'08: Proceedings of the 2008 International Congress on Image and Signal Processing. Piscataway: IEEE, 2008, 5: 441-446.
[11] DIELEMAN S, SCHRAUWEN B. Multiscale approaches to music audio feature learning [C]//ISMIR 2013: Proceedings of the 14th International Society for Music Information Retrieval Conference. Rio de Janeiro: Pontifícia Universidade Católica do Paraná, 2013: 116-121.
[12] ZENG C, XU J. Notice of retraction multi-modal music genre classification approach [C]//ICCSIT 2010: Proceeding of 2010 3rd IEEE International Conference on Computer Science and Information Technology. Piscataway: IEEE, 2010, 8: 398-402.
[13] MUSIC TYPES [EB/OL]. [2014-12-03]. http://baike.baidu.com/view/157245.htm.(音乐类型[EB/OL].[2014-12-03].http://baike.baidu.com/view/157245.htm.)

[1]	范黎林, 曹富康, 王琬婷, 杨凯, 宋钊瑜. 基于需求模式自适应匹配的间歇性需求预测方法[J]. 《计算机应用》唯一官方网站, 2024, 44(9): 2747-2755.
[2]	陈虹, 齐兵, 金海波, 武聪, 张立昂. 融合1D-CNN与BiGRU的类不平衡流量异常检测[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2493-2499.
[3]	雷明珠, 王浩, 贾蓉, 白琳, 潘晓英. 基于特征间关系合成少数类样本的过采样算法[J]. 《计算机应用》唯一官方网站, 2024, 44(5): 1428-1436.
[4]	高麟, 周宇, 邝得互. 进化双层自适应局部特征选择[J]. 《计算机应用》唯一官方网站, 2024, 44(5): 1408-1414.
[5]	孙敏, 成倩, 丁希宁. 基于CBAM-CGRU-SVM的Android恶意软件检测方法[J]. 《计算机应用》唯一官方网站, 2024, 44(5): 1539-1545.
[6]	孙林, 刘梦含. 基于自适应布谷鸟优化特征选择的K-means聚类[J]. 《计算机应用》唯一官方网站, 2024, 44(3): 831-841.
[7]	徐大鹏, 侯新民. 基于网络结构设计的图神经网络特征选择方法[J]. 《计算机应用》唯一官方网站, 2024, 44(3): 663-670.
[8]	孟圣洁, 于万钧, 陈颖. 最大相关和最大差异的高维数据特征选择算法[J]. 《计算机应用》唯一官方网站, 2024, 44(3): 767-771.
[9]	刘晶鑫, 黄雯静, 徐亮胜, 黄冲, 吴建生. 字典学习与样本关联保持结合的无监督特征选择模型[J]. 《计算机应用》唯一官方网站, 2024, 44(12): 3766-3775.
[10]	乔恩保, 高向阳, 程俊. 基于支持向量机的自恢复自适应蒙特卡洛定位算法[J]. 《计算机应用》唯一官方网站, 2024, 44(10): 3246-3251.
[11]	何添, 沈宗鑫, 黄倩倩, 黄雁勇. 基于自适应学习的多视图无监督特征选择方法[J]. 《计算机应用》唯一官方网站, 2023, 43(9): 2657-2664.
[12]	黄学雨, 贺怀宇, 林慧敏, 陈金水. 基于特征聚合的铜合金金相图分类识别方法[J]. 《计算机应用》唯一官方网站, 2023, 43(8): 2593-2601.
[13]	孙林, 黄金旭, 徐久成. 基于邻域容差互信息和鲸鱼优化算法的非平衡数据特征选择[J]. 《计算机应用》唯一官方网站, 2023, 43(6): 1842-1854.
[14]	于振华, 刘争气, 刘颖, 郭城. 基于自适应混合粒子群优化的软件缺陷预测特征选择方法[J]. 《计算机应用》唯一官方网站, 2023, 43(4): 1206-1213.
[15]	陈容均, 严宣辉, 杨超城. 面向时间序列的混合图像化循环胶囊分类网络[J]. 《计算机应用》唯一官方网站, 2023, 43(3): 692-699.

基于多核学习支持向量机的音乐流派分类

Music genre classification based on multiple kernel learning and support vector machine

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics