Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Capacity measurement of a recurrent inhibitory neural network

Capacity measurement of a recurrent inhibitory neural network Yuan and Leibold BMC Neuroscience 2011, 12(Suppl 1):P196 http://www.biomedcentral.com/1471-2202/12/S1/P196 POSTER PRESENTATION Open Access Capacity measurement of a recurrent inhibitory neural network Chun-Wei Yuan , Christian Leibold From Twentieth Annual Computational Neuroscience Meeting: CNS*2011 Stockholm, Sweden. 23-28 July 2011 References Inhibitory neurons are considered to play a central role 1. Maass W, Natschläger T, Markram H: Real-time computing without stable as rhythm generator and in shaping feed-forward recep- states: a new framework for neural computation based on tive fields. While much attention has been paid to such perturbations. Neural Comput 2002, 14:2531-2560. 2. Jäger H, Haas H: Harnessing nonlinearity: predicting chaotic systems and effectson excitatoryneurons,littleis donetostudy saving energy in wireless communication. Science 2004, 304:78-80. theseinhibitoryneurons ’ ability to directly process doi:10.1186/1471-2202-12-S1-P196 information. Here we present a model that investigates Cite this article as: Yuan and Leibold: Capacity measurement of a the computational capacity of a recurrent inhibitory recurrent inhibitory neural network. BMC Neuroscience 2011 12(Suppl 1): neural network. P196. Our work focuses on quantifying the performance of a recurrent network of inhibitory integrate-and-fire neu- rons in canonical classification tasks. The model begins with parallel independent excitatory Poisson inputs con- nected to the recurrent network. Then, the network out- put is feed-forwardly directed to a read-out linear classifier. An identical network, but with zero synaptic connectivity, is set up for benchmarking. The analysis is then conducted by comparing the capacities of both set- ups, at 95% accuracy, as a function of parameters such as inhibitory weight, network size, etc. It is found that, in general, neurons with faster time constants provide better computational power. Further- more, there is an optimum weight amongst the inhibi- tory neurons that yields at least a 20% network performance improvement (Figure 1). The inhibition plays the role of suppressing overdriven, stereotypical firing behavior to render efficient sparse encoding of temporal information. This illustrates that the nonli- nearity of a recurrent, dynamical network possesses Figure 1 Classification accuracy vs. number of training patterns, for more computational capacity than a simple feed-forward a fully connected inhibitory network of N = 100 neurons. Performance of the zero-connectivity network is shown as a linear expansion provided by the non-connected net- benchmark for comparison. The N-dimensional network output is work [1,2]. binned into n patterns, with a bin size of 30 ms, and the linear classifier is trained to separate the first n/2 patterns from the latter n/2. The input weight, with exponentially decreasing post-synaptic Published: 18 July 2011 current (psc, with 100 ms time constant), is sub-threshold. Inhibitory network weights are reported in relation to the input weight, * Correspondence: yuan@bio.lmu.de though the inhibitory psc time constants are much smaller (8 ms). Division of Neurobiology, Department of Biology II, Ludwig Maximilians Universität, 82152 Martinsried, Germany Full list of author information is available at the end of the article © 2011 Yuan and Leibold; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BMC Neuroscience Springer Journals

Capacity measurement of a recurrent inhibitory neural network

Yuan, Chun-Wei; Leibold, Christian
BMC Neuroscience , Volume 12 (1) – Jul 18, 2011
Download PDF
1 page

Loading next page...
 
/lp/springer-journals/capacity-measurement-of-a-recurrent-inhibitory-neural-network-4e6q00mHDx

References (2)

Publisher
Springer Journals
Copyright
Copyright © 2011 by Yuan and Leibold; licensee BioMed Central Ltd.
Subject
Biomedicine; Neurosciences; Neurobiology; Animal Models
eISSN
1471-2202
DOI
10.1186/1471-2202-12-S1-P196
Publisher site
See Article on Publisher Site

Abstract

Yuan and Leibold BMC Neuroscience 2011, 12(Suppl 1):P196 http://www.biomedcentral.com/1471-2202/12/S1/P196 POSTER PRESENTATION Open Access Capacity measurement of a recurrent inhibitory neural network Chun-Wei Yuan , Christian Leibold From Twentieth Annual Computational Neuroscience Meeting: CNS*2011 Stockholm, Sweden. 23-28 July 2011 References Inhibitory neurons are considered to play a central role 1. Maass W, Natschläger T, Markram H: Real-time computing without stable as rhythm generator and in shaping feed-forward recep- states: a new framework for neural computation based on tive fields. While much attention has been paid to such perturbations. Neural Comput 2002, 14:2531-2560. 2. Jäger H, Haas H: Harnessing nonlinearity: predicting chaotic systems and effectson excitatoryneurons,littleis donetostudy saving energy in wireless communication. Science 2004, 304:78-80. theseinhibitoryneurons ’ ability to directly process doi:10.1186/1471-2202-12-S1-P196 information. Here we present a model that investigates Cite this article as: Yuan and Leibold: Capacity measurement of a the computational capacity of a recurrent inhibitory recurrent inhibitory neural network. BMC Neuroscience 2011 12(Suppl 1): neural network. P196. Our work focuses on quantifying the performance of a recurrent network of inhibitory integrate-and-fire neu- rons in canonical classification tasks. The model begins with parallel independent excitatory Poisson inputs con- nected to the recurrent network. Then, the network out- put is feed-forwardly directed to a read-out linear classifier. An identical network, but with zero synaptic connectivity, is set up for benchmarking. The analysis is then conducted by comparing the capacities of both set- ups, at 95% accuracy, as a function of parameters such as inhibitory weight, network size, etc. It is found that, in general, neurons with faster time constants provide better computational power. Further- more, there is an optimum weight amongst the inhibi- tory neurons that yields at least a 20% network performance improvement (Figure 1). The inhibition plays the role of suppressing overdriven, stereotypical firing behavior to render efficient sparse encoding of temporal information. This illustrates that the nonli- nearity of a recurrent, dynamical network possesses Figure 1 Classification accuracy vs. number of training patterns, for more computational capacity than a simple feed-forward a fully connected inhibitory network of N = 100 neurons. Performance of the zero-connectivity network is shown as a linear expansion provided by the non-connected net- benchmark for comparison. The N-dimensional network output is work [1,2]. binned into n patterns, with a bin size of 30 ms, and the linear classifier is trained to separate the first n/2 patterns from the latter n/2. The input weight, with exponentially decreasing post-synaptic Published: 18 July 2011 current (psc, with 100 ms time constant), is sub-threshold. Inhibitory network weights are reported in relation to the input weight, * Correspondence: yuan@bio.lmu.de though the inhibitory psc time constants are much smaller (8 ms). Division of Neurobiology, Department of Biology II, Ludwig Maximilians Universität, 82152 Martinsried, Germany Full list of author information is available at the end of the article © 2011 Yuan and Leibold; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Journal

BMC Neuroscience – Springer Journals

Published: Jul 18, 2011

There are no references for this article.