Regensburg 2019 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

BP: Fachverband Biologische Physik

BP 12: Poster II

BP 12.56: Poster

Dienstag, 2. April 2019, 14:00–16:00, Poster B2

Estimating autocorrelation times of subsampled autoregressive processes under non-stationary parameters — •Jorge de Heuvel, Jens Wilting, Viola Priesemann, Johannes Zierenberg, and Paul Spitzner — Max-Planck-Institut für Dynamik und Selbstorganisation, Göttingen, Deutschland

Collective cortical dynamics are often approximated by linear autoregressive models. Only recent methodological advances enable us to infer network autocorrelation times even under subsampling, assuming stationarity. However, cortical dynamics are likely subject to non-stationary input, which can lead to a severe overestimation of the network autocorrelation time. Here, we present a novel approach for the subsampling-invariant estimation of network autocorrelation even under non-stationary input. A trial based experimental setup is applied, where the external input is time-dependent but similar over each trial. Our estimator is verified on both numerical data and experimental recordings.