Published 2020 | Version v1
Book

Spiking network model with STDP learning and inhibitory interneuronal connections for solving a classification task

  • 1. Natsional'nyj Issledovatel'skij Yadernyj Univ. «MIFI», Moscow (Russian Federation)
  • 2. Natsional'nyj Issledovatel'skij Tsentr «Kurchatovskij Inst.», Moscow (Russian Federation)

Description

For a neural network comprising one excitatory layer and one inhibitory layer, the biologically-inspired Spike-Timing-Dependent Plasticity is shown to be capable of learning not only the image recognition task of MNIST as known previously, but also a real-valued classification task of Fisher's Iris

Abstract (Russian)

Для сети из одного слоя возбуждающих и одного слоя тормозящих нейронов показано, что на основе биологически мотивированной модели долговременной синаптической пластичности Spike-Timing-Dependent Plasticity возможно обучение сети не только распознаванию изображений, что было ранее показано на примере рукописных цифр MNIST, но и классификации векторов рациональных чисел, на примере задачи ирисов Фишера
Part of:
VI International conference «Laser, plasma research and technologies – LaPlaz-2020». Collection of scientific papers. Part 1

Additional details

Additional titles

Original title (Russian)
Модел' спайковой нейронной сети с STDP-обучением и тормозящими межнейронными связями для решения задачи классификации

Publishing Information

Publisher
NIYaU MIFI
Imprint Place
Moscow (Russian Federation)
ISBN
978-5-7262-2655-2
Imprint Title
VI International conference #Left-Pointing Double Angle Quotation Mark#Laser, plasma research and technologies #En Dash# LaPlaz-2020#Right-Pointing Double Angle Quotation Mark#. Collection of scientific papers. Part 1
Imprint Pagination
463 p.
Journal Page Range
p. 103-104

Conference

Title
International Conference on Laser, Plasma Research and Technologies
Original Conference Title
VI Mezhdunarodnaya konferentsiya «Lazernye, plazmennye issledovaniya i tekhnologii – LaPlaz-2020»
Acronym
LaPlas 2020
Dates
11-14 Feb 2020
Place
Moscow (Russian Federation)

INIS

Country of Publication
Russian Federation
Country of Input or Organization
Russian Federation
INIS RN
53056534
Subject category
S97: MATHEMATICAL METHODS AND COMPUTING;
Resource subtype / Literary indicator
Conference
Descriptors DEI
ADAPTIVE SYSTEMS; ALGORITHMS; ARTIFICIAL INTELLIGENCE; NEURAL NETWORKS; PATTERN RECOGNITION
Descriptors DEC
COMPUTERIZED CONTROL SYSTEMS; CONTROL SYSTEMS; MATHEMATICAL LOGIC; ON-LINE CONTROL SYSTEMS; ON-LINE SYSTEMS

Optional Information

Notes
2 refs. Imprint:VI Mezhdunarodnaya konferentsiya #Left-Pointing Double Angle Quotation Mark#Lazernye, plazmennye issledovaniya i tekhnologii #En Dash# LaPlaz-2020#Right-Pointing Double Angle Quotation Mark#. Sbornik nauchnykh trudov. Chast' 1