Performance Assessment of Smart Electricity Meters Based on Max Margin Bayesian Classifiers
There are large amount of quality and monitoring data of electricity meters, it is crucial and valuable to assess the operating performance of the smart electricity meters automatically with this data. In this paper, we propose an intelligent operating performance assessment method for smart electric meters based on selective ensemble of max margin Bayesian classifiers. The genetic algorithm is firstly used to select most relevant attributes for assessment, and max margin Bayesian classifiers is utilized to make the assessment. We use the bagging and clustering to ensemble multiple classifiers to obtain better results. The experimental results illustrate the efficiency and effectiveness of the proposed method.
KeywordsEnsemble learning Smart electricity meter Performance assessment
- 1.Zhu, Y., Qing, X., Liu, J., Tian, Z., Zhou, C.: Data mining application in smart electric meter fault analysis. Jiangsu Electr. Eng. 35(05), 19–23 (2016)Google Scholar
- 2.Xu, D., Luan, W., Wang, P., Zhang, Z.: Analysis and applications of smart electric meter data. Distrib. Utilization 32(8), 25–30 (2015)Google Scholar
- 4.Guo, Y., Wilkinson, D., Schuurmans, D.: Maximum margin Bayesian networks. In: Proceedings International Conference on Uncertainty in Artificial Intelligence, pp. 233–242 (2005)Google Scholar
- 5.Sha, F., Saul, L.: Comparison of large margin training to other discriminative methods for phonetic recognition by hidden Markov models. In: Proceedings IEEE International Conference on Acoustics, Speech and Signal Processing, pp. 313–316 (2007)Google Scholar
- 9.Liu, D., Zhu, Y., Ni, N., Liu, J.: Ordered proposition fusion based on consistency and uncertainty measurements. Sci. China Inf. Sci. 60(8), 1–19 (2017). 082103Google Scholar