THE DISPLAY OF PROBABILISTIC SOLUTIONS IN SONAR TARGET CLASSIFICATION

Abstract

The comparative meaningfulness of five, alter native methods of displaying classification in formation was investigated. Three displays were based on likelihood ratio, a fourth was a display of 20 solution lights such as that used with MITEC, and the fifth was based on the traditional Navy three- category classification (trichotomy). The three displays based on likelihood ratio produced interpretations very similar to those of a theoretically ideal interpreter. There was some unwillingness to extract the appropriate amount of certainty from very small or very large ratios. The 20-lights display was interpreted essentially as a linear, equal-interval scale of probability, an interpretation not intended by its designers. Overly conservative reactions, in relation to an ideal interpreter, occurred both for moderately low and moderately high probabilities. The traditional Navy trichotomy produced severely limited interpretations, the phrase 'possible submarine' producing reactions of nearly maximum uncertainty. There were substantial differences in interpretation among the subjects. The problem of intelligent design of probabilistic displays is discussed along with the need of the designer of clue correlating de vices to match the output display with the inter pretations of potential users.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 01, 1963
Accession Number
AD0335934

Entities

People

  • Robert R. Mackie

Tags

Communities of Interest

  • C4I
  • Ground and Sea Platforms
  • Human Systems
  • Sensors
  • Weapons Technologies

DTIC Thesaurus Topics

  • Acoustics
  • Active Sonar
  • Antisubmarine Warfare
  • Computer Programming
  • Detection
  • Human Factors Engineering
  • Naval Operations
  • Navy
  • Observers
  • Processing Equipment
  • Ships
  • Sonar Targets
  • Submarines
  • Target Classification
  • Target Echoes
  • Test And Evaluation
  • Training

Readers

  • Computational Linguistics
  • Computational Modeling and Simulation
  • Educational Psychology