TY  - JOUR
AU  - Nestler, Sandra
AU  - Helias, Moritz
AU  - Gilson, Matthieu
TI  - Statistical temporal pattern extraction by neuronal architecture
JO  - Physical review research
VL  - 5
IS  - 3
SN  - 2643-1564
CY  - College Park, MD
PB  - APS
M1  - FZJ-2023-03196
SP  - 033177
PY  - 2023
AB  - Neuronal systems need to process temporal signals. Here, we show how higher-order temporal (co)fluctuationscan be employed to represent and process information. Concretely, we demonstrate that a simple biologicallyinspired feedforward neuronal model can extract information from up to the third-order cumulant to performtime series classification. This model relies on a weighted linear summation of synaptic inputs followed bya nonlinear gain function. Training both the synaptic weights and the nonlinear gain function exposes how thenonlinearity allows for the transfer of higher-order correlations to the mean, which in turn enables the synergisticuse of information encoded in multiple cumulants to maximize the classification accuracy. The approach isdemonstrated both on synthetic and real-world datasets of multivariate time series. Moreover, we show thatthe biologically inspired architecture makes better use of the number of trainable parameters than a classicalmachine-learning scheme. Our findings emphasize the benefit of biological neuronal architectures, paired withdedicated learning algorithms, for the processing of information embedded in higher-order statistical cumulantsof temporal (co)fluctuations.
LB  - PUB:(DE-HGF)16
UR  - <Go to ISI:>//WOS:001074650500002
DO  - DOI:10.1103/PhysRevResearch.5.033177
UR  - https://juser.fz-juelich.de/record/1010690
ER  -