公开(公告)号：US20180253640A1

公开(公告)日：2018-09-06

申请号：US15909918

申请日：2018-03-01

Applicant: STC.UNM

Inventor： Alireza Goudarzi ,  Darko Stefanovic ,  Steven Wayde Graves ,  Daniel Kalb

IPC: G06N3/04 ,  G06N3/08

CPC classification number: G06N3/0454 ,  G06N3/0445 ,  G06N3/088 ,  G16H50/00

Abstract: The invention is directed to a hybrid architecture that comprises a stacked autoencoder and a deep echo state layer for temporal pattern discovery in high-dimensional sequence data. The stacked autoencoder plays a preprocessing role that exploits spatial structure in data and creates a compact representation. The compact representation is then fed to the echo state layer in order to generate a short-term memory of the inputs. The output of the network may be trained to generate any target output.

公开(公告)号：US11188813B2

公开(公告)日：2021-11-30

申请号：US15909918

申请日：2018-03-01

Applicant: STC.UNM

Inventor： Alireza Goudarzi ,  Darko Stefanovic ,  Steven Wayde Graves ,  Daniel Kalb

IPC: G06N3/04 ,  G06N3/08 ,  G16H50/20

Abstract: The invention is directed to a hybrid architecture that comprises a stacked autoencoder and a deep echo state layer for temporal pattern discovery in high-dimensional sequence data. The stacked autoencoder plays a preprocessing role that exploits spatial structure in data and creates a compact representation. The compact representation is then fed to the echo state layer in order to generate a short-term memory of the inputs. The output of the network may be trained to generate any target output.