A Mixed Regression Model for the Analysis of Explosives Performance Data.

Abstract

An intra-class regression model with additive error terms consisting of a gage calibration error, a within-class performance variation, and a general experimental error is developed for the analysis of explosives performance data associated with the shock waves of underwater explosions. Parameter estimates for this mixed model are sought by means of the maximum likelihood and restricted maximum likelihood techniques. Derivatives of the likelihood function suitable for constrained Newton-Raphson optimization are derived. Useful applications are discussed.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jul 30, 1982
Accession Number
ADA125337

Entities

People

  • William W. Mcdonald

Organizations

  • Naval Ordnance Laboratory

Tags

Communities of Interest

  • Counter IED
  • Weapons Technologies

DTIC Thesaurus Topics

  • Algorithms
  • Computational Science
  • Computations
  • Confidence Limits
  • Data Mining
  • Data Science
  • Explosives
  • Information Processing
  • Information Science
  • Military Research
  • Operations Research
  • Plastic Explosives
  • Random Variables
  • Regression Analysis
  • Statistical Algorithms
  • Statistical Analysis
  • Surveys

Fields of Study

  • Mathematics

Readers

  • Explosive Engineering.
  • Operations Research
  • Regression Analysis.