Distributed Parameter Estimation via Pseudo-likelihood
This addresses the need for efficient distributed learning in sensor networks, offering a practical solution with incremental improvements over existing methods.
The paper tackles the problem of distributed parameter estimation in sensor networks by proposing a general approach based on combining local pseudo-likelihood estimators, showing that simple methods like linear combination or max-voting with second-order information are statistically competitive with more advanced joint optimization, with low communication and computational costs.
Estimating statistical models within sensor networks requires distributed algorithms, in which both data and computation are distributed across the nodes of the network. We propose a general approach for distributed learning based on combining local estimators defined by pseudo-likelihood components, encompassing a number of combination methods, and provide both theoretical and experimental analysis. We show that simple linear combination or max-voting methods, when combined with second-order information, are statistically competitive with more advanced and costly joint optimization. Our algorithms have many attractive properties including low communication and computational cost and "any-time" behavior.