OPTIMAL COMBINATION POLICIES FOR ADAPTIVE SOCIAL LEARNING

被引：3

作者：

Hu, Ping ^{[1
]}

Bordignon, Virginia ^{[1
]}

Vlaski, Stefan ^{[2
]}

Saye, Ali H. ^{[1
]}

机构：

[1] Ecole Polytech Fed Lausanne EPFL, Sch Engn, Lausanne, Switzerland

[2] Imperial Coll London, Dept Elect & Elect Engn, London, England

来源：

2022 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP) | 2022年

关键词：

Social learning; combination policy; large deviations; adaptation time; NETWORKS;

D O I：

10.1109/ICASSP43922.2022.9746784

中图分类号：

O42 [声学];

学科分类号：

070206 ; 082403 ;

摘要：

This paper investigates the effect of combination policies on the performance of adaptive social learning in non-stationary environments. By analyzing the relation between the error probability and the underlying graph topology, we prove that in the slow adaptation regime, combination policies with a uniform Perron eigenvector will provide the smallest steady-state error probability. This result indicates that in terms of learning accuracy, doubly-stochastic combination policies yield optimal performance. Moreover, we estimate the adaptation time of adaptive social learning in the small signal-to-noise regime and show that in this regime, the influence of combination policies on the adaptation time is insignificant.

引用

页码：5842 / 5846

页数：5

共 20 条

[1] Adaptive Social Learning [J].

Bordignon, Virginia ;

Matta, Vincenzo ;

Sayed, Ali H. .

IEEE TRANSACTIONS ON INFORMATION THEORY, 2021, 67 (09) :6053-6081

[2] Asymptotic Optimality of Running Consensus in Testing Binary Hypotheses [J].

Braca, Paolo ;

Marano, Stefano ;

Matta, Vincenzo ;

Willett, Peter .

IEEE TRANSACTIONS ON SIGNAL PROCESSING, 2010, 58 (02) :814-825

[3]

Den Hollander Frank, 2008, Large deviations, V14

[4]

Jadbabaie A., 2013, Paper 13-28

[5] Non-Bayesian social learning [J].

Jadbabaie, Ali ;

Molavi, Pooya ;

Sandroni, Alvaro ;

Tahbaz-Salehi, Alireza .

GAMES AND ECONOMIC BEHAVIOR, 2012, 76 (01) :210-225

[6]

Kassam S.A., 1987, Signal Detection in Non-Gaussian Noise, V1

[7] Social Learning and Distributed Hypothesis Testing [J].

Lalitha, Anusha ;

Javidi, Tara ;

Sarwate, Anand D. .

IEEE TRANSACTIONS ON INFORMATION THEORY, 2018, 64 (09) :6161-6179

[8] Interplay Between Topology and Social Learning Over Weak Graphs [J].

Matta, Vincenzo ;

Bordignon, Virginia ;

Santos, Augusto ;

Sayed, Ali H. .

IEEE OPEN JOURNAL OF SIGNAL PROCESSING, 2020, 1 :99-119

[9] Distributed Detection Over Adaptive Networks: Refined Asymptotics and the Role of Connectivity [J].

Matta, Vincenzo ;

Braca, Paolo ;

Marano, Stefano ;

Sayed, Ali H. .

IEEE TRANSACTIONS ON SIGNAL AND INFORMATION PROCESSING OVER NETWORKS, 2016, 2 (04) :442-460

[10] Diffusion-Based Adaptive Distributed Detection: Steady-State Performance in the Slow Adaptation Regime [J].

Matta, Vincenzo ;

Braca, Paolo ;

Marano, Stefano ;

Sayed, Ali H. .

IEEE TRANSACTIONS ON INFORMATION THEORY, 2016, 62 (08) :4710-4732

← 1 2 →