Showing 1–1 of 1 results for author: Kanold, P

Efficient Estimation of Compressible State-Space Models with Application to Calcium Signal Deconvolution

Authors: Abbas Kazemipour, Ji Liu, Patrick Kanold, Min Wu, Behtash Babadi

Abstract: In this paper, we consider linear state-space models with compressible innovations and convergent transition matrices in order to model spatiotemporally sparse transient events. We perform parameter and state estimation using a dynamic compressed sensing framework and develop an efficient solution consisting of two nested Expectation-Maximization (EM) algorithms. Under suitable sparsity assumption… ▽ More In this paper, we consider linear state-space models with compressible innovations and convergent transition matrices in order to model spatiotemporally sparse transient events. We perform parameter and state estimation using a dynamic compressed sensing framework and develop an efficient solution consisting of two nested Expectation-Maximization (EM) algorithms. Under suitable sparsity assumptions on the innovations, we prove recovery guarantees and derive confidence bounds for the state estimates. We provide simulation studies as well as application to spike deconvolution from calcium imaging data which verify our theoretical results and show significant improvement over existing algorithms. △ Less

Submitted 20 October, 2016; originally announced October 2016.

Comments: 2016 IEEE Global Conference on Signal and Information Processing (GlobalSIP), Dec. 7-9, 2016, Washington D.C