A Distributed Learning Based on Robust Diffusion Sgd Over Adaptive Networks with Noisy Output Data

作     者：Barani, Fatemeh Savadi, Abdorreza Yazdi, Hadi Sadoghi 

作者机构： Mashhad Iran Center of Excellence on Soft Computing and Intelligent Information Processing Mashhad Iran 

出 版 物：《SSRN》 

年 卷 期：2023年

核心收录：

主　　题：Diffusion 

摘      要：Outliers and noises are unavoidable factors that cause the performance of the distributed learning algorithms to be severely reduced. Developing a robust algorithm is vital in applications such as system identification and forecasting stock market, in which noise on the desired signals may intensely divert the solutions. In this paper, we propose a Robust Diffusion SGD (RDSGD) algorithm based on the pseudo-Huber objective function which can significantly suppress the effect of Gaussian and non-Gaussian noises on estimation performances in the adaptive networks. Performance and convergence behavior of RDSGD are assessed in presence of the α-stable and Mixed-Gaussian noises in the stationary and non-stationary environments. Simulation results show that the proposed algorithm can achieve both higher convergence rate and lower steady-state misadjustment than the conventional diffusion algorithms and several robust algorithms in the presence of non-Gaussian noises. On the other hand, when the environment noise is Gaussian, the performance of RDSGD is similar to the diffusion SGD algorithm with the square loss function (DSGD), while outperforms the other mentioned algorithms. © 2023, The Authors. All rights reserved.