Analyzing a Method to Determine the Utility of Adding a Classification System to a Sequence for Improved Accuracy

Abstract

Frequently, ensembles of classification systems are combined into a sequence in order to better enhance the accuracy in classifying objects of interest. However, there is a point in which adding an additional system to a sequence no longer enhances the sequence. Either the increase in operational costs exceeds the improvement in accuracy or the addition of the system does not increase accuracy at all. This research will examine a utility measure that determines the valid or invalid nature of adding a classification system to a sequence based on the ratio of the change in accuracy to the change in operational costs. Three sequential strategies defined on a two-class population outcome will be examined: Believe the Positive, Believe the Negative, and Believe the Extreme. This work expands upon known accuracy and cost equations for each strategy in order to generalize them for any fixed sequence length. Through simulation, this research identified which characteristics have the greatest impact on the utility measure and provides guidance on the threshold value of the utility measure that differentiates between when the addition of a system to the sequence may be useful (valid) and when it is not (invalid).

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 21, 2019
Accession Number
AD1075550

Entities

People

  • Kevin S. Pamilagas

Organizations

  • Air Force Institute of Technology

Tags

Communities of Interest

  • Human Systems

DTIC Thesaurus Topics

  • Accuracy
  • Air Force
  • Applied Mathematics
  • Classification
  • Covariance
  • Distribution Functions
  • Equations
  • Factor Analysis
  • Governments
  • Information Science
  • Mathematics
  • Probability
  • Sequences
  • Simulations
  • Standards
  • United States
  • United States Government

Readers

  • Computational Modeling and Simulation
  • Mathematical Modeling and Probability Theory.
  • Regression Analysis.