多特征融合和机器学习算法的电子音乐分类模型  

Classification Model of Electronic Music Based on Multi Feature Fusion and Machine Learning Algorithm

作　　者：易伶 YI Ling(Art College, Shangluo University, Shangluo 726000, China)

机构地区：[1]商洛学院艺术学院,陕西商洛726000

出　　处：《微型电脑应用》2020年第9期117-119,共3页Microcomputer Applications

摘　　要：电子音乐分类有利于电子音乐的在线检索,当前电子音乐分类模型难以准确识别各种类型的电子音乐,使得当前电子音乐分类模型分类效果差,为了提高电子音乐分类正确率,提出了多特征融合和机器学习算法的电子音乐分类模型。首先采集电子音乐信号,并从电子音乐信号中提取分类的多种特征,然后采用机器学习算法描述电子音乐信号类型和特征之间的联系,建立电子音乐分类器,最后采用模型对多种电子音乐进行分类仿真实验,结果表明,相对于当前其它电子音乐分类模型,该模型减少了电子音乐分类器构建的时间,加快了电子音乐分类速度,能够高精度识别各种类型的电子音乐,电子音乐分类正确率明显提高,验证了该电子音乐分类模型的优越性。Classification of electronic music is conducive to online retrieval of electronic music.The current classification model of electronic music is difficult to accurately identify various types of electronic music,which makes the classification effect of current classification model of electronic music poor.In order to improve the accuracy of classification of electronic music,a classification model of electronic music based on multi feature fusion and machine learning algorithm is proposed.Firstly,the electronic music signals are collected,and the classification features are extracted from the electronic music signals.Then,the machine learning algorithm is used to describe the relationship between the types and features of the electronic music signals,and the electronic music classifier is established.Finally,the simulation experiment of this paper model is used to classify the electronic music signals.The results show that compared with other current electronic music classification models,this model reduces the time of constructing the electronic music classifier,speeds up the classification speed of electronic music,and can recognize all kinds of electronic music with high precision.The accuracy of electronic music classification is obviously improved,which verifies the superiority of this model.

关 键 词：电子音乐 机器学习算法 分类特征 分类器设计 训练时间 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]