沈然,王庆娟,金良峰,丁麒
SHEN Ran,WANG Qingjuan,JIN Liangfeng,DING Qi

• 1:国网浙江省电力有限公司营销服务中心

摘要(Abstract)：

自动识别电网客户语音中的情感对电网运营具有重要的意义，语音情感分类模型往往需要专家根据经验设计，这限制了人工智能算法在这一模型上的应用。提出了一种基于自动机器学习的电网客户语音情感分类算法。首先设计了神经网络搜索空间构建方法，构建了基础的神经网络基本组成单元，并搭建了对应的元架构空间。然后，为了提高情感分类模型的准确率，利用强化学习进行神经网络模型和超参数的求解。实验结果表明，该算法可以有效完成电网客户情感分类任务，情感分类准确率达到90.93%，而且整个过程不依赖于人工设计，具有较高的效率和准确率。
The recognition of emotion in speech of customers is of great significance to grid operation. However,the recognition model of emotion in speech is designed by experts,which imposes restrictions on the application of artificial intelligence. This paper proposes a classification algorithm of emotion in speech for power grid customers. The algorithm designs a neural network search space construction method,and constructs the basic components of the neural network and the corresponding meta-architecture space. Then,in order to improve the performance of the final emotion recognition model,reinforcement learning is used to solve the neural network model and hyperparameters. Experimental results show that the algorithm can effectively complete the recognition of emotion in speech,and the accuracy of classification accuracy is up to 90.93%;moreover,the whole process does not rely on manual design,and has high efficiency and accuracy.

关键词(KeyWords)： 自动机器学习;情感识别;搜索空间构建;强化学习
AutoML;emotion classification;search space construction;reinforcement learning

Abstract:

Keywords:

基金项目(Foundation): 国网浙江省电力有限公司科技项目（5211YF200052）

作者(Author): 沈然,王庆娟,金良峰,丁麒
SHEN Ran,WANG Qingjuan,JIN Liangfeng,DING Qi

DOI: 10.19585/j.zjdl.202205012

参考文献(References)：

• [1]韩国华，丁永刚，尹凤武.基于大数据的电力营销管理创新分析[J].电力设备管理，2021(6):130-132.
• [2]王楚，王忠锋，徐志远，等.电力营业厅智能服务机器人个性化情感分析技术研究[J].供用电，2020,37(6):21-26.
• [3] SAXENA A,KHANNA A,GUPTA D.Emotion recognition and detection methods:a comprehensive survey[J].Journal of Artificial Intelligence and Systems,2020,2(1):53-79.
• [4] SAILUNAZ K,DHALIWAL M,ROKNE J,et al.Emotion detection from text and speech:a survey[J]. Social Network Analysis and Mining,2018,8(1):1-26.
• [5] XIONG Z Y,RADHAKRISHNAN R,DIVAKARAN A,et al. Comparing MFCC and MPEG-7 audio features for feature extraction,maximum likelihood HMM and entropic prior HMM for sports audio classification[C]//Proceedings of IEEE International Conference on Acoustics,Speech,and Signal Processing,July 6-9,2003,Baltimore,MD,USA:III-397.
• [6] ZHU H,LUO M D,WANG R,et al.Deep audio-visual learning:a survey[J].International Journal of Automation and Computing,2021,18(3):351-376.
• [7] PURWINS H,LI B,VIRTANEN T,et al.Deep learning for audio signal processing[J]. IEEE Journal of Selected Topics in Signal Processing,2019,13(2):206-219.
• [8] Z?LLER M A,HUBER M F.Benchmark and survey of automated machine learning frameworks[J].Journal of Artificial Intelligence Research,2021,70:409-472.
• [9] BARREIRO E,MUNTEANU C R,GESTAL M,et al.Net-net AutoML selection of artificial neural network topology for brain connectome prediction[J]. Applied Sciences,2020,10(4):1308.
• [10] VACCARO L,SANSONETTI G,MICARELLI A.An empirical review of automated machine learning[J].Computers,2021,10(1):11.
• [11] LIU C X,CHEN L C,SCHROFF F,et al. AutoDeepLab:hierarchical neural architecture search for semantic image segmentation[C]//2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR),Long Beach,CA,USA:2019:82-92.
• [12] ELSKEN T,METZEN J H,HUTTER F.Neural architecture search:a survey[J].The Journal of Machine Learning Research,2019,20(1):1997-2017.
• [13] JIN H F,SONG Q Q,HU X.Auto-keras:an efficient neural architecture search system[C]//KDD′19:Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery&Data Mining.2019:1946-1956.
• [14] JAAFRA Y,LUC LAURENT J,DERUYVER A,et al.Reinforcement learning for neural architecture search:a review[J].Image and Vision Computing,2019,89:57-66.
• [15] HENDERSON P,ISLAM R,BACHMAN P,et al.Deep reinforcement learning that matters[C]//Proceedings of the AAAI Conference on Artificial Intelligence,April 29,2018.
• [16] ZOPH B,VASUDEVAN V,SHLENS J,et al.Learning transferable architectures for scalable image recognition[J]. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition,2018:8697-8710.
• [17] SHERSTINSKY A.Fundamentals of recurrent neural network(RNN)and long short-term memory(LSTM)network[J]. Physica D:Nonlinear Phenomena,2020,404:132306.
• [18] HERSHEY S,CHAUDHURI S,ELLIS D P W,et al.CNN architectures for large-scale audio classification[C]//2017 IEEE International Conference on Acoustics,Speech and Signal Processing. New Orleans,LA,USA.IEEE,2017:131-135.
• [19] SANTURKAR S,TSIPRAS D,ILYAS A,et al. How does batch normalization help optimization?[C]//Proceedings of the 32nd International Conference on Neural Information Processing Systems,December 3,2018:2488-2498.
• [20]张小飞，张道银，郑珞琳，等.基于机器学习算法的电力信息网络安全态势感知研究[J].电器与能效管理技术，2021(8):16-23.
• [21]杨金成，郭泽林，袁铁江，等.大数据环境下基于改进SVM的典型负荷类型识别[J].电力电容器与无功补偿，2021,42(4):170-175.
• [22]周楠，张平，郑征，等.基于机器学习的电力通信网带宽分配算法[J].电网与清洁能源，2021,37(5):67-73.
• [23]陆春阳，郭天驹，李晓溪.基于大数据的事故舆情分析系统研究[J].电力大数据，2021,24(1):70-78.
• [24]王蕊，俞凌枫，朱斌，等.大数据和机器学习技术在电缆运维中的应用研究[J].电器与能效管理技术，2020(12):114-118.
• [25]雒军，唐坚，赵喆，等.基于大数据的电力环保数据平台建设[J].电力大数据，2020,23(1):58-63.