Learning Probabilistic Inference through Spike-Timing-Dependent Plasticity

Dejan Pecevski, Wolfgang Maass

Research output: Contribution to journalArticleResearchpeer-review

Abstract

Numerous experimental data show that the brain is able to extract information from complex, uncertain, and often ambiguous experiences. Furthermore, it can use such learnt information for decision making through probabilistic inference. Several models have been proposed that aim at explaining how probabilistic inference could be performed by networks of neurons in the brain. We propose here a model that can also explain how such neural network could acquire the necessary information for that from examples. We show that spike-timing-dependent plasticity in combination with intrinsic plasticity generates in ensembles of pyramidal cells with lateral inhibition a fundamental building block for that: probabilistic associations between neurons that represent through their firing current values of random variables. Furthermore, by combining such adaptive network motifs in a recursive manner the resulting network is enabled to extract statistical information from complex input streams, and to build an internal model for the distribution p (*) that generates the examples it receives. This holds even if p (*) contains higher-order moments. The analysis of this learning process is supported by a rigorous theoretical foundation. Furthermore, we show that the network can use the learnt internal model immediately for prediction, decision making, and other types of probabilistic inference.

Original languageEnglish
JournaleNeuro
Volume3
Issue number2
DOIs
Publication statusPublished - 16 Jul 2016

Fingerprint

Decision Making
Learning
Neurons
Pyramidal Cells
Brain
Complex Mixtures
Inhibition (Psychology)

Keywords

  • Action Potentials
  • Brain
  • Computer Simulation
  • Humans
  • Models, Neurological
  • Nerve Net
  • Neuronal Plasticity
  • Neurons
  • Probability Learning
  • Time Factors
  • Journal Article
  • Research Support, Non-U.S. Gov't

Cite this

Learning Probabilistic Inference through Spike-Timing-Dependent Plasticity. / Pecevski, Dejan; Maass, Wolfgang.

In: eNeuro, Vol. 3, No. 2, 16.07.2016.

Research output: Contribution to journalArticleResearchpeer-review

@article{b3f22a3e021c49c59ae341a94d2e083e,
title = "Learning Probabilistic Inference through Spike-Timing-Dependent Plasticity",
abstract = "Numerous experimental data show that the brain is able to extract information from complex, uncertain, and often ambiguous experiences. Furthermore, it can use such learnt information for decision making through probabilistic inference. Several models have been proposed that aim at explaining how probabilistic inference could be performed by networks of neurons in the brain. We propose here a model that can also explain how such neural network could acquire the necessary information for that from examples. We show that spike-timing-dependent plasticity in combination with intrinsic plasticity generates in ensembles of pyramidal cells with lateral inhibition a fundamental building block for that: probabilistic associations between neurons that represent through their firing current values of random variables. Furthermore, by combining such adaptive network motifs in a recursive manner the resulting network is enabled to extract statistical information from complex input streams, and to build an internal model for the distribution p (*) that generates the examples it receives. This holds even if p (*) contains higher-order moments. The analysis of this learning process is supported by a rigorous theoretical foundation. Furthermore, we show that the network can use the learnt internal model immediately for prediction, decision making, and other types of probabilistic inference.",
keywords = "Action Potentials, Brain, Computer Simulation, Humans, Models, Neurological, Nerve Net, Neuronal Plasticity, Neurons, Probability Learning, Time Factors, Journal Article, Research Support, Non-U.S. Gov't",
author = "Dejan Pecevski and Wolfgang Maass",
year = "2016",
month = "7",
day = "16",
doi = "10.1523/ENEURO.0048-15.2016",
language = "English",
volume = "3",
journal = "eNeuro",
issn = "2373-2822",
publisher = "Society for Neuroscience",
number = "2",

}

TY - JOUR

T1 - Learning Probabilistic Inference through Spike-Timing-Dependent Plasticity

AU - Pecevski, Dejan

AU - Maass, Wolfgang

PY - 2016/7/16

Y1 - 2016/7/16

N2 - Numerous experimental data show that the brain is able to extract information from complex, uncertain, and often ambiguous experiences. Furthermore, it can use such learnt information for decision making through probabilistic inference. Several models have been proposed that aim at explaining how probabilistic inference could be performed by networks of neurons in the brain. We propose here a model that can also explain how such neural network could acquire the necessary information for that from examples. We show that spike-timing-dependent plasticity in combination with intrinsic plasticity generates in ensembles of pyramidal cells with lateral inhibition a fundamental building block for that: probabilistic associations between neurons that represent through their firing current values of random variables. Furthermore, by combining such adaptive network motifs in a recursive manner the resulting network is enabled to extract statistical information from complex input streams, and to build an internal model for the distribution p (*) that generates the examples it receives. This holds even if p (*) contains higher-order moments. The analysis of this learning process is supported by a rigorous theoretical foundation. Furthermore, we show that the network can use the learnt internal model immediately for prediction, decision making, and other types of probabilistic inference.

AB - Numerous experimental data show that the brain is able to extract information from complex, uncertain, and often ambiguous experiences. Furthermore, it can use such learnt information for decision making through probabilistic inference. Several models have been proposed that aim at explaining how probabilistic inference could be performed by networks of neurons in the brain. We propose here a model that can also explain how such neural network could acquire the necessary information for that from examples. We show that spike-timing-dependent plasticity in combination with intrinsic plasticity generates in ensembles of pyramidal cells with lateral inhibition a fundamental building block for that: probabilistic associations between neurons that represent through their firing current values of random variables. Furthermore, by combining such adaptive network motifs in a recursive manner the resulting network is enabled to extract statistical information from complex input streams, and to build an internal model for the distribution p (*) that generates the examples it receives. This holds even if p (*) contains higher-order moments. The analysis of this learning process is supported by a rigorous theoretical foundation. Furthermore, we show that the network can use the learnt internal model immediately for prediction, decision making, and other types of probabilistic inference.

KW - Action Potentials

KW - Brain

KW - Computer Simulation

KW - Humans

KW - Models, Neurological

KW - Nerve Net

KW - Neuronal Plasticity

KW - Neurons

KW - Probability Learning

KW - Time Factors

KW - Journal Article

KW - Research Support, Non-U.S. Gov't

U2 - 10.1523/ENEURO.0048-15.2016

DO - 10.1523/ENEURO.0048-15.2016

M3 - Article

VL - 3

JO - eNeuro

JF - eNeuro

SN - 2373-2822

IS - 2

ER -