Article

Boolean factor analysis by attractor neural network.

Institute of Higher Nervous Activity and Neurophysiology, Russian Academy of Science, Moscow 119991, Russia.
IEEE Transactions on Neural Networks (impact factor: 2.95). 06/2007; 18(3):698-707. DOI:10.1109/TNN.2007.891664 pp.698-707
Source: PubMed

ABSTRACT A common problem encountered in disciplines such as statistics, data analysis, signal processing, textual data representation, and neural network research, is finding a suitable representation of the data in the lower dimension space. One of the principles used for this reason is a factor analysis. In this paper, we show that Hebbian learning and a Hopfield-like neural network could be used for a natural procedure for Boolean factor analysis. To ensure efficient Boolean factor analysis, we propose our original modification not only of Hopfield network architecture but also its dynamics as well. In this paper, we describe neural network implementation of the Boolean factor analysis method. We show the advantages of our Hopfield-like network modification step by step on artificially generated data. At the end, we show the efficiency of the method on artificial data containing a known list of factors. Our approach has the advantage of being able to analyze very large data sets while preserving the nature of the data.

0 0
 · 
2 Bookmarks
 · 
42 Views
  • Source
    Article: Analysis of continuous attractors for 2-D linear threshold neural networks.
    Lan Zou, Huajin Tang, Kay Chen Tan, Weinian Zhang
    [show abstract] [hide abstract]
    ABSTRACT: This brief investigates continuous attractors of the well-developed model in visual cortex, i.e., the linear threshold (LT) neural networks, based on a parameterized 2-D model. On the basis of existing results on nondegenerate equilibria in mathematics, we further discuss degenerate equilibria for such networks and present properties and distributions of the equilibria, which enables us to draw the coexistence conditions of nondegenerate and degenerate equilibria (e.g., singular lines). Our theoretical results provide a useful framework for precise tuning on the network parameters, e.g., the feedbacks and visual inputs. Simulations are also presented to illustrate the theoretical findings.
    IEEE Transactions on Neural Networks 02/2009; 20(1):175-80. · 2.95 Impact Factor
  • Source
    Article: Binary Inference for Primary User Separation in Cognitive Radio Networks
    Huy Nguyen, Guanbo Zheng, Zhu Han, Rong Zheng
    [show abstract] [hide abstract]
    ABSTRACT: Spectrum sensing receives much attention recently in the cognitive radio (CR) network research, i.e., secondary users (SUs) constantly monitor channel condition to detect the presence of the primary users (PUs). In this paper, we go beyond spectrum sensing and introduce the PU separation problem, which concerns with the issues of distinguishing and characterizing PUs in the context of collaborative spectrum sensing and monitor selection. The observations of monitors are modeled as boolean OR mixtures of underlying binary sources for PUs. We first justify the use of the binary OR mixture model as opposed to the traditional linear mixture model through simulation studies. Then we devise a novel binary inference algorithm for PU separation. Not only PU-SU relationship are revealed, but PUs' transmission statistics and activities at each time slot can also be inferred. Simulation results show that without any prior knowledge regarding PUs' activities, the algorithm achieves high inference accuracy even in the presence of noisy measurements. Comment: Submitted to the IEEE Transactions on Wireless Communications on Nov. 2010
    12/2010;
  • Source
    Article: Methods for structuring scientific knowledge from many areas related to aging research.
    Alex Zhavoronkov, Charles R Cantor
    [show abstract] [hide abstract]
    ABSTRACT: Aging and age-related disease represents a substantial quantity of current natural, social and behavioral science research efforts. Presently, no centralized system exists for tracking aging research projects across numerous research disciplines. The multidisciplinary nature of this research complicates the understanding of underlying project categories, the establishment of project relations, and the development of a unified project classification scheme. We have developed a highly visual database, the International Aging Research Portfolio (IARP), available at AgingPortfolio.org to address this issue. The database integrates information on research grants, peer-reviewed publications, and issued patent applications from multiple sources. Additionally, the database uses flexible project classification mechanisms and tools for analyzing project associations and trends. This system enables scientists to search the centralized project database, to classify and categorize aging projects, and to analyze the funding aspects across multiple research disciplines. The IARP is designed to provide improved allocation and prioritization of scarce research funding, to reduce project overlap and improve scientific collaboration thereby accelerating scientific and medical progress in a rapidly growing area of research. Grant applications often precede publications and some grants do not result in publications, thus, this system provides utility to investigate an earlier and broader view on research activity in many research disciplines. This project is a first attempt to provide a centralized database system for research grants and to categorize aging research projects into multiple subcategories utilizing both advanced machine algorithms and a hierarchical environment for scientific collaboration.
    PLoS ONE 01/2011; 6(7):e22597. · 4.09 Impact Factor

Show more

Data provided are for informational purposes only. Although carefully collected, accuracy cannot be guaranteed. The impact factor represents a rough estimation of the journal's impact factor and does not reflect the actual current impact factor. Publisher conditions are provided by RoMEO. Differing provisions from the publisher's actual policy or licence agreement may be applicable.

Keywords

advantages
 
artificial data
 
Boolean factor analysis
 
Boolean factor analysis method
 
data analysis
 
efficient Boolean factor analysis
 
factor analysis
 
factors
 
Hebbian
 
Hopfield network architecture
 
Hopfield-like network modification step
 
Hopfield-like neural network
 
known list
 
large data sets
 
lower dimension space
 
neural network implementation
 
neural network research
 
original modification
 
suitable representation
 
textual data representation