标题：Component Pursuit from Noisy Measurements
作者：Zhao, Yongjian ;Jiang, Bin
作者机构：[Zhao, Yongjian ;Jiang, Bin ] Shandong University, Weihai, Shandong; 264209, China
会议名称：2018 International Conference on Smart Grid and Electrical Automation, ICSGEA 2018
会议日期：9 June 2018 through 10 June 2018
来源：Proceedings - 2018 International Conference on Smart Grid and Electrical Automation, ICSGEA 2018
关键词：Component; Noise; Property; Pursuit; Statistics; Variable
摘要：To achieve efficient component pursuit from noisy measurements, a learning algorithm is presented that combines standard gradient principle and the standard stochastic approximations. By extending the linear predictor principle from noise-free case, a proper objective function is introduced which has the same generic form as that for the noise-free case. Extensive computer simulations are performed to illustrate the power of the presented technique. © 2018 IEEE.