Statistical and Modeling Uncertainties in the Thermal Target Acquisition Model Improvement Program (TAMIP) Predictions.

Abstract

We analyze the uncenainties that are associated with the Thermal Target Acquisition Model Improvement Program (TAMIP) target detection predictions. The deviations of the measured probabilities from the predicted ones are much smaller than for the previous model, but still exceed those that would be expected from the finite statistical samples. Therefore there is still some residual error in the prediction. We determine confidence limits on the predictor variable which are unbiased in the sense that they are accurate at both high and low detection probabilities. We validate the confidence interval using a second data set that was not used in the development of this phase of the target detection model.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 1995
Accession Number
ADA301182

Entities

People

  • James D. Silk

Organizations

  • Institute for Defense Analyses

Tags

DTIC Thesaurus Topics

  • Acquisition
  • Confidence Limits
  • Data Science
  • Data Sets
  • Detection
  • Detectors
  • Information Science
  • Intervals
  • Probability
  • Residuals
  • Statistical Samples
  • Target Acquisition
  • Target Detection
  • Targets
  • Thermal Targets
  • Uncertainty

Readers

  • Computational Modeling and Simulation
  • Image Processing and Computer Vision.
  • Statistical inference.