Analysis of Naval Aviation Selection Test Data with Nonlinear Models. Part 1. Parameter Estimation.

Abstract

The purpose of this paper is basically tutorial in nature and, as such, describes an algorithm for estimating the parameters of a nonlinear model. This algorithm is called 'simulated annealing.' The actual workings of this algorithm are examined in some detail. The reason for studying this algorithm is because statistical analysis of naval aviation selection test data has always relied on the use of linear regression models. Linear models represent only a small subset of possible mathematical models that could be used as an empirical tool to predict aviator performance. Specifically, the whole class of nonlinear models has not been addressed. Recent research into neural networks and parallel distributed processing has uncovered some interesting nonlinear models. We intend to reanalyze the test scores of student naval aviators with a nonlinear model borrowed from the neural network literature. We hope that this new class of nonlinear models will be a more powerful tool in predicting aviator performance and will result in an improved naval aviator selection test battery.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 01, 1990
Accession Number
ADA238244

Entities

People

  • David J. Blower

Organizations

  • Naval Aerospace Medical Research Laboratory

Tags

DTIC Thesaurus Topics

  • Algorithms
  • Annealing
  • Artificial Intelligence
  • Bayesian Networks
  • Computer Programs
  • Computer Simulations
  • Computers
  • Data Science
  • Ground State
  • High Temperature
  • Information Science
  • Mathematical Models
  • Monte Carlo Method
  • Naval Aviation
  • Neural Networks
  • Probabilistic Models
  • Statistical Analysis

Readers

  • Adaptive Control and Estimation with Uncertainty in Dynamic Systems.
  • Aviation Science / Aeronautics.
  • Theoretical Analysis.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Machine Learning Algorithms