Gaussian mixture clustering algorithm combining elbow method and expectation-maximization for power system customer segmentation

doi:10.11772/j.issn.1001-9081.2020050672

Journal of Computer Applications ›› 2020, Vol. 40 ›› Issue (11): 3217-3223.DOI: 10.11772/j.issn.1001-9081.2020050672

• Data science and technology • Previous Articles Next Articles

Gaussian mixture clustering algorithm combining elbow method and expectation-maximization for power system customer segmentation

CHEN Yu¹, TIAN Bojin², PENG Yunzhu², LIAO Yong³

1. Information Communication Branch Company, Chongqing Guanghui Power Supply Service Company Limited, Chongqing 400014, China;
2. Information Communication Branch Company, State Grid Chongqing Electric Power Company, Chongqing 401120, China;
3. School of Microelectronics and Communication Engineering, Chongqing University, Chongqing 400444, China

Received:2020-05-20 Revised:2020-07-22 Online:2020-11-10 Published:2020-08-14
Supported by:
This work is partially supported by the Science and Technology Project of State Grid Chongqing Electric Power Company (2019 Yudian Technology 29#).

联合手肘法和期望最大化的高斯混合聚类电力系统客户分群算法

陈聿¹, 田博今², 彭云竹², 廖勇³

1. 重庆广汇供电服务有限责任公司信息通信分公司, 重庆 400014;
2. 国网重庆市电力公司信息通信分公司, 重庆 401120;
3. 重庆大学微电子与通信工程学院, 重庆 400044

通讯作者: 廖勇(1982-),男,四川自贡人,副研究员,博士,CCF杰出会员,主要研究方向:移动通信、人工智能;liaoy@cqu.edu.cn
作者简介:陈聿(1983-),女,上海人,工程师,主要研究方向:通信工程;田博今(1985-),男,重庆人,工程师,硕士,主要研究方向:电力信息通信运行;彭云竹(1992-),重庆人,女,工程师,主要研究方向:信息技术、项目管理
基金资助:
国网重庆市电力公司科学技术项目（2019渝电科技29#）。

Abstract

Abstract: In order to further improve the user experience of power system customers, and aiming at the problems of poor optimization ability, lack of compactness and difficulty in solving the optimal number of clusters, a Gaussian mixture clustering algorithm combining elbow method and Expectation-Maximization (EM) was proposed, which can mine the potential information in a large number of customer data. The good clustering results were obtained by EM algorithm iteration. Aiming at the shortcoming of the traditional Gaussian mixture clustering algorithm that needs to obtain the number of user clusters in advance, the number of customer clusters was reasonably found by using elbow method. The case study shows that compared with hierarchical clustering algorithm and K-Means algorithm, the proposed algorithm has the increase of both FM (Fowlkes-Mallows) and AR (Adjusted-Rand) indexes more than 10%, and the decrease of Compactness Index (CI) and Degree of Separation (DS) less than 15% and 25% respectively. It can be seen that the performance of the algorithm is greatly improved.

Key words: power system, customer segmentation, Gaussian mixture model clustering, accurate service, Expectation-Maximization (EM), elbow method

摘要： 为进一步提升电力系统客户的用户体验，针对现有聚类算法寻优能力差、紧凑性不足以及较难求解聚类数目最优值的问题，提出一种联合手肘法与期望最大化（EM）的高斯混合聚类算法，挖掘大量客户数据中的潜在信息。该算法通过EM算法迭代出良好的聚类结果，而针对传统的高斯混合聚类算法需要提前获取用户分群数量的缺点，利用手肘法合理找出客户的分群数量。案例分析表明，所提算法与层次聚类算法和K-Means算法相比，FM、AR指标的增幅均超过10%，紧凑度（CI）和分离度（DS）的降幅分别低于15%和25%，可见性能有较大提升。

关键词: 电力系统, 客户分群, 高斯混合模型聚类, 精准服务, 期望最大化, 手肘法

CLC Number:

TP391

CHEN Yu, TIAN Bojin, PENG Yunzhu, LIAO Yong. Gaussian mixture clustering algorithm combining elbow method and expectation-maximization for power system customer segmentation[J]. Journal of Computer Applications, 2020, 40(11): 3217-3223.

陈聿, 田博今, 彭云竹, 廖勇. 联合手肘法和期望最大化的高斯混合聚类电力系统客户分群算法[J]. 计算机应用, 2020, 40(11): 3217-3223.

References

[1] 全海金, 何映思. 基于大数据的改进模糊K-means算法[J]. 重庆理工大学学报(自然科学版),2018,32(12):145-148. (QUAN H J,HE Y S. Improved fuzzy K-means algorithm based on big data[J]. Journal of Chongqing University of Technology (Natural Science Edition),2018,32(12):145-148.)
[2] 吴会会, 高淑萍, 彭弘铭, 等. 自适应模糊C均值聚类的数据融合算法[J]. 计算机工程与应用,2019, 55(5):326-35. (WU H H, GAO S P,PENG H M,et al. Data fusion algorithm for adaptive fuzzy C-means clustering[J]. Computer Engineering and Applications,2019,55(5):26-35.)
[3] 任莎莎, 郎文辉. 基于K-GMM算法的SAR海冰图像分类[J]. 地理与地理信息科学, 2018,34(5):42-48. (REN S S,LANG W H. SAR sea ice image classification based on K-GMM algorithm[J]. Geography and Geo-Information Science,2018,34(5):42-48.)
[4] SONG M,WANG H. Highly efficient incremental estimation of Gaussian mixture models for online data stream clustering[C]//Proceedings of the SPIE 5803,Intelligent Computing:Theory and Applications Ⅲ. Bellingham,WA:SPIE,2005:No. 601724.
[5] SYAKUR M A, KHOTIMAH B K, ROCHMAN S, et al. Integration k-means clustering method and elbow method for identification of the best customer profile cluster[J]. IOP Conference Series:Materials Science and Engineering,2018,336:No. 12017.
[6] 吴广建, 章剑林, 袁丁. 基于K-means的手肘法自动获取K值方法研究[J]. 软件, 2019,40(5):167-170. (WU G J,ZHANG J L, YUAN D. Automatically obtaining K value based on K-mean elbow method[J]. Computer Engineering and Software,2019,40(5):167-170.)
[7] REYNOLDS D A, QUATIERI T F, DUNN R B. Speaker verification using adapted gaussian mixture models[J]. Digital Signal Processing,2000,10(1/2/3):19-41.
[8] MOON T K. The expectation-maximization algorithm[J]. IEEE Signal Processing Magazine,1996,13(6):47-60.
[9] BERTSEKAS D P. Constrained optimization and Lagrange multiplier methods[M]. Washington:Academic Press, 1982:95-15.
[10] 傅立伟, 武森. 基于属性值集中度的分类数据聚类有效性内部评价指标[J]. 工程科学学报, 2019,41(5):682-693. (FU L W, WU S. Internal evaluation index of classification data clustering effectiveness based on attribute value concentration[J]. Journal of Engineering Sciences,2019,41(5):682-693.)
[11] 谢娟英. 无监督学习方法及其应用[M]. 北京:电子工业出版社,2016:72-86. (XIE J Y. Unsupervised Learning Methods and Applications[M]. Beijing:Publishing House of Electronic Industry,2016:72-86.)
[12] XIE J Y,GAO H C,XIE W X,et al. Robust clustering by detecting density peaks and assigning points based on fuzzy weighted K-nearest neighbors[J]. Information Sciences,2016, 354:19-40.
[13] KAPP A V,TIBSHIRANI R. Are clusters found in one dataset present in another dataset?[J]. Biostatistics,2007,8(1):9-31.
[14] DAVIES D L,BOULDIN D W. A cluster separation measure[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1979,PAMI-1(2):224-227.
[15] HASHIMOTO W,NAKAMURA T,MIYAMOTO S. Comparison and evaluation of different cluster validity measures including their kernelization[J]. Journal of Advanced Computational Intelligence and Intelligent Informatics,2009,13(3):204-209.
[16] 谢娟英, 周颖, 王明钊, 等. 聚类有效性评价新指标[J]. 智能系统学报, 2017,12(6):873-882. (XIE J Y,ZHOU Y,WANG M Z,et al. New criteria for evaluating the validity of clustering[J]. CAAI Transactions on Intelligent Systems, 2017, 12(6):873-882.)
[17] SANGAM R S,OM H. The k-modes algorithm with entropy based similarity coefficient[J]. Procedia Computer Science,2015,50:93-98.

Gaussian mixture clustering algorithm combining elbow method and expectation-maximization for power system customer segmentation

联合手肘法和期望最大化的高斯混合聚类电力系统客户分群算法

PDF

Knowledge

Abstract

Cite this article

share this article

References

Related Articles 14

Recommended Articles

Metrics

[1]	WANG Yang, WU Fan, YAO Zongqiang, LIU Jie, LI Dong. Residential electricity consumption analysis based on regularized matrix factorization [J]. Journal of Computer Applications, 2017, 37(8): 2405-2409.
[2]	YU Qun, ZHANG Min, CAO Na, HE Qing, SHI Liang. Power grid fault evolution model based on fuzzy cellular automata [J]. Journal of Computer Applications, 2015, 35(9): 2682-2686.
[3]	QIAN Shuqu, WU Huihong, XU Guofeng. Constrained multiobjective optimization algorithm based on repairing strategy for solving dynamic environment/economic dispatch [J]. Journal of Computer Applications, 2015, 35(8): 2249-2255.
[4]	HUANG Shunliang WANG Qi. Method for customer segmentation based on three-way decisions theory [J]. Journal of Computer Applications, 2014, 34(1): 244-248.
[5]	HAO Zedong YU Songsong GUAN Jihong. Hyperspectral image classification based on active learning [J]. Journal of Computer Applications, 2013, 33(12): 3441-3443.
[6]	ZHANG Wei HUANG Wei XIA Limin. Recommendation research based on general content probabilistic latent semantic analysis model [J]. Journal of Computer Applications, 2013, 33(05): 1330-1333.
[7]	YAO Xin-an SONG Fei HU Shi-ping. Design of power supply for high-performance computer system [J]. Journal of Computer Applications, 2012, 32(04): 1176-1179.
[8]	QIN Jun XIAO Rong. Clustering user behavior patterns of E-commerce search engine based on mixture of Markov models [J]. Journal of Computer Applications, 2012, 32(04): 1086-1089.
[9]	LONG Yin YIN Heng-jing ZHU Jiang LI Fang-wei. New adaptive dynamic channel allocation algorithm in cognitive Radio [J]. Journal of Computer Applications, 2011, 31(11): 2915-2917.
[10]	CHENG Yanwei XIE Yongcheng LI Guangsheng. Vehicle power system condition prediction based on weighted hidden Markov model [J]. Journal of Computer Applications, 2011, 31(06): 1696-1698.
[11]	HUANG Haifeng ZHANG Keheng ZHANG Hong JI Xuechun CHEN Peng. Key technologies of dynamic information database for power systems [J]. Journal of Computer Applications, 2011, 31(06): 1681-1684.
[12]	. Empirical research on customer segmentation of securities based on clustering [J]. Journal of Computer Applications, 2010, 30(2): 495-498.
[13]	. Clustering algorithm based on complex attributes similarity and its applications [J]. Journal of Computer Applications, 2010, 30(07): 1930-1932.
[14]	LIU Qing-shan, JIANG Xiao-hua(. Distributed control scheme for power transformation system based on network agent of VxWorks [J]. Journal of Computer Applications, 2005, 25(02): 433-436.