Efficient Bayesian spatial prediction with mobile sensor networks using Gaussian Markov random fields

Authors:
Yunfei Xu;Jongeun Choi;Sarat Dass;Tapabrata Maiti
Affiliations:
-;-;-;-
Venue:
Automatica (Journal of IFAC)
Year:
2013

Citing 12
Cited 0

Parallel and distributed computation: numerical methods

Parallel and distributed computation: numerical methods
A Bayesian Committee Machine

Neural Computation
Brief paper: Extremum seeking for moderately unstable systems and for autonomous vehicle target tracking without position measurements

Automatica (Journal of IFAC)
Conditional-mean least-squares fitting of Gaussian Markov random fields to Gaussian fields

Computational Statistics & Data Analysis
Fast kriging of large data sets with Gaussian Markov random fields

Computational Statistics & Data Analysis
Near-Optimal Sensor Placements in Gaussian Processes: Theory, Efficient Algorithms and Empirical Studies

The Journal of Machine Learning Research
Distributed learning and cooperative control for multi-agent systems

Automatica (Journal of IFAC)
Extremum seeking under stochastic noise and applications to mobile sensors

Automatica (Journal of IFAC)
Decentralized Environmental Modeling by Mobile Sensor Networks

IEEE Transactions on Robotics
Mobile Sensor Network Navigation Using Gaussian Processes With Truncated Observations

IEEE Transactions on Robotics
Spatial prediction with mobile sensor networks using Gaussian processes with built-in Gaussian Markov random fields

Automatica (Journal of IFAC)
Distributed parametric and nonparametric regression with on-line performance bounds computation

Automatica (Journal of IFAC)

Quantified Score

Hi-index	22.14

Visualization

Abstract

In this paper, we consider the problem of predicting a large scale spatial field using successive noisy measurements obtained by mobile sensing agents. The physical spatial field of interest is discretized and modeled by a Gaussian Markov random field (GMRF) with uncertain hyperparameters. From a Bayesian perspective, we design a sequential prediction algorithm to exactly compute the predictive inference of the random field. The main advantages of the proposed algorithm are: (1) the computational efficiency due to the sparse structure of the precision matrix, and (2) the scalability as the number of measurements increases. Thus, the prediction algorithm correctly takes into account the uncertainty in hyperparameters in a Bayesian way and is also scalable to be usable for mobile sensor networks with limited resources. We also present a distributed version of the prediction algorithm for a special case. An adaptive sampling strategy is presented for mobile sensing agents to find the most informative locations in taking future measurements in order to minimize the prediction error and the uncertainty in hyperparameters simultaneously. The effectiveness of the proposed algorithms is illustrated by numerical experiments.