System and Method for Feature Set Reduction.

Abstract

A system and method for ranking features by exploiting their relationship to the Fisher projection space. The system ranks n features in a feature set using a design set comprising exemplars from each of M possible event classes of an associated feature based classification system. A training set is created by randomly selecting exemplars from each of the M classes in the design set. A "smoothed" Fisher projection space for the training set is created by replacing the sample means and the within class sample covariance matrix normally used in deriving a Fisher projection space with expressions for the mean vectors and covariance matrices derived from event class probability density function estimates. The angle between a given feature and the smoothed Fisher projection space is calculated for each feature in the feature set, and the features are then ordered by increasing numerical size of this angle. The system produces a reduced feature set by eliminating those features which are not important for classification based on the linear ranking of the features.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jun 30, 1995
Accession Number
ADD017757

Entities

People

  • Roy L. Streit
  • Stephen G. Greineder
  • Tod Luginbuhl

Organizations

  • United States Department of the Navy

Tags

Communities of Interest

  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Compilers
  • Computers
  • Data Science
  • Eigenvalues
  • Eigenvectors
  • Extraction
  • Factor Analysis
  • Feature Extraction
  • Generators
  • Information Science
  • Inventions
  • Pattern Recognition
  • Probability
  • Probability Density Functions
  • Random Variables
  • Recognition
  • Test And Evaluation

Readers

  • Adaptive Control and Estimation with Uncertainty in Dynamic Systems.
  • Approximation Theory.
  • Neural Network Machine Learning.

Technology Areas

  • Space