Linear Regression to a Lower Order Model: Effects and Implications

Abstract

Linear regression is used extensively in the fields of science, engineering, and business. Because data-gathering processes can be complex, show random effects, or be unknown, often times the regression model used only approximates the actual process model. This report analyzes the effects of using a reduced-order process model in linear regression. In particular, the relationship of the Taylor series coefficients to the regression parameters is discussed. Relationships are generated to equate regression and Taylor series parameters, and an error analysis is performed to compare the effects of noise and modeling errors.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Nov 15, 1989
Accession Number
ADA218258

Entities

People

  • M. L. Graham

Organizations

  • Naval Undersea Warfare Center

Tags

Communities of Interest

  • Ground and Sea Platforms

DTIC Thesaurus Topics

  • Coefficients
  • Compensation
  • Control Systems
  • Data Rate
  • Engineering
  • Equations
  • Errors
  • Gaussian Noise
  • Interdisciplinary Science
  • Intervals
  • Low Noise
  • Measurement
  • Military Research
  • Noise
  • Rhode Island
  • Standards
  • Time Intervals

Readers

  • Adaptive Control and Estimation with Uncertainty in Dynamic Systems.
  • Computational Modeling and Simulation
  • Regression Analysis.