Distributed multivariate regression with unknown noise covariance in the presence of outliers: an MDL approach
Document typeConference report
PublisherInstitute of Electrical and Electronics Engineers (IEEE)
Rights accessOpen Access
We consider the problem of estimating the coefficients in a multivariable linear model by means of a wireless sensor network which may be affected by anomalous measurements. The noise covariance matrices at the different sensors are assumed unknown. Treating outlying samples, and their support, as additional nuisance parameters, the Maximum Likelihood estimate is investigated, with the number of outliers being estimated according to the Minimum Description Length principle. A distributed implementation based on iterative consensus techniques is then proposed, and it is shown effective for managing outliers in the data.
CitationLópez, R., Romero, D., Sala, J., Pages, A. Distributed multivariate regression with unknown noise covariance in the presence of outliers: an MDL approach. A: IEEE Statistical Signal Processing Workshop. "2016 IEEE Statistical Signal Processing Workshop (SSP) took place 25-29 June 2016 in Palma de Mallorca, Spain". Palma de Mallorca: Institute of Electrical and Electronics Engineers (IEEE), 2016.