skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Visual Empirical Region of Influence (VERI) Pattern Recognition Algorithms

Abstract

We developed new pattern recognition (PR) algorithms based on a human visual perception model. We named these algorithms Visual Empirical Region of Influence (VERI) algorithms. To compare the new algorithm's effectiveness against othe PR algorithms, we benchmarked their clustering capabilities with a standard set of two-dimensional data that is well known in the PR community. The VERI algorithm succeeded in clustering all the data correctly. No existing algorithm had previously clustered all the pattens in the data set successfully. The commands to execute VERI algorithms are quite difficult to master when executed from a DOS command line. The algorithm requires several parameters to operate correctly. From our own experiences we realized that if we wanted to provide a new data analysis tool to the PR community we would have to provide a new data analysis tool to the PR community we would have to make the tool powerful, yet easy and intuitive to use. That was our motivation for developing graphical user interfaces (GUI's) to the VERI algorithms. We developed GUI's to control the VERI algorithm in a single pass mode and in an optimization mode. We also developed a visualization technique that allows users to graphically animate and visuallymore » inspect multi-dimensional data after it has been classified by the VERI algorithms. The visualization technique that allows users to graphically animate and visually inspect multi-dimensional data after it has been classified by the VERI algorithms. The visualization package is integrated into the single pass interface. Both the single pass interface and optimization interface are part of the PR software package we have developed and make available to other users. The single pass mode only finds PR results for the sets of features in the data set that are manually requested by the user. The optimization model uses a brute force method of searching through the cominations of features in a data set for features that produce the best pattern recognition results. With a small number of features in a data set an exact solution can be determined. However, the number of possible combinations increases exponentially with the number of features and an alternate means of finding a solution must be found. We developed and implemented a technique for finding solutions in data sets with both small and large numbers of features. The VERI interface tools were written using the Tcl/Tk GUI programming language, version 8.1. Although the Tcl/Tk packages are designed to run on multiple computer platforms, we have concentrated our efforts to develop a user interface for the ubiquitous DOS environment. The VERI algorithms are compiled, executable programs. The interfaces run the VERI algorithms in Leave-One-Out mode using the Euclidean metric.« less

Authors:
; ;
Publication Date:
Research Org.:
Sandia National Laboratories (SNL), Albuquerque, NM, and Livermore, CA (United States)
Sponsoring Org.:
USDOE
OSTI Identifier:
1230544
Report Number(s):
VERI-PRAV2.4; 001493IBMPC00
DOE Contract Number:  
DE-AC04-94AL85000
Resource Type:
Software
Software Revision:
00
Software Package Number:
001493
Software Package Contents:
Media Directory; Software Abstract; Media includes Source Code; Executable Module(s); Sample Problem Input Data; Installation Instructions; User Guide; and Sample Problem Output Data / 1 CD ROM
Software CPU:
IBMPC
Open Source:
No
Source Code Available:
Yes
Related Software:
Tcl/Tk 8.1 software packages
Country of Publication:
United States

Citation Formats

Osboum, Gordon C., Martinez, Rubel F., and Bartholomew, John W. Visual Empirical Region of Influence (VERI) Pattern Recognition Algorithms. Computer software. Vers. 00. USDOE. 1 May. 2002. Web.
Osboum, Gordon C., Martinez, Rubel F., & Bartholomew, John W. (2002, May 1). Visual Empirical Region of Influence (VERI) Pattern Recognition Algorithms (Version 00) [Computer software].
Osboum, Gordon C., Martinez, Rubel F., and Bartholomew, John W. Visual Empirical Region of Influence (VERI) Pattern Recognition Algorithms. Computer software. Version 00. May 1, 2002.
@misc{osti_1230544,
title = {Visual Empirical Region of Influence (VERI) Pattern Recognition Algorithms, Version 00},
author = {Osboum, Gordon C. and Martinez, Rubel F. and Bartholomew, John W.},
abstractNote = {We developed new pattern recognition (PR) algorithms based on a human visual perception model. We named these algorithms Visual Empirical Region of Influence (VERI) algorithms. To compare the new algorithm's effectiveness against othe PR algorithms, we benchmarked their clustering capabilities with a standard set of two-dimensional data that is well known in the PR community. The VERI algorithm succeeded in clustering all the data correctly. No existing algorithm had previously clustered all the pattens in the data set successfully. The commands to execute VERI algorithms are quite difficult to master when executed from a DOS command line. The algorithm requires several parameters to operate correctly. From our own experiences we realized that if we wanted to provide a new data analysis tool to the PR community we would have to provide a new data analysis tool to the PR community we would have to make the tool powerful, yet easy and intuitive to use. That was our motivation for developing graphical user interfaces (GUI's) to the VERI algorithms. We developed GUI's to control the VERI algorithm in a single pass mode and in an optimization mode. We also developed a visualization technique that allows users to graphically animate and visually inspect multi-dimensional data after it has been classified by the VERI algorithms. The visualization technique that allows users to graphically animate and visually inspect multi-dimensional data after it has been classified by the VERI algorithms. The visualization package is integrated into the single pass interface. Both the single pass interface and optimization interface are part of the PR software package we have developed and make available to other users. The single pass mode only finds PR results for the sets of features in the data set that are manually requested by the user. The optimization model uses a brute force method of searching through the cominations of features in a data set for features that produce the best pattern recognition results. With a small number of features in a data set an exact solution can be determined. However, the number of possible combinations increases exponentially with the number of features and an alternate means of finding a solution must be found. We developed and implemented a technique for finding solutions in data sets with both small and large numbers of features. The VERI interface tools were written using the Tcl/Tk GUI programming language, version 8.1. Although the Tcl/Tk packages are designed to run on multiple computer platforms, we have concentrated our efforts to develop a user interface for the ubiquitous DOS environment. The VERI algorithms are compiled, executable programs. The interfaces run the VERI algorithms in Leave-One-Out mode using the Euclidean metric.},
doi = {},
url = {https://www.osti.gov/biblio/1230544}, year = {Wed May 01 00:00:00 EDT 2002},
month = {Wed May 01 00:00:00 EDT 2002},
note =
}

Software:
To order this software, request consultation services, or receive further information, please fill out the following request.

Save / Share:

To receive further information, fill out the request form below. OSTI staff will begin to process an order for scientific and technical software once the signed site license agreement is received. You may also reach us by email at: .

Software Request

(required)
(required)
(required)
(required)
(required)
(required)
(required)
(required)