Date of Award

3-2004

Document Type

Thesis

Degree Name

Master of Science

Department

Department of Operational Sciences

First Advisor

John O. Miller, PhD

Abstract

Uncertainty is an age old element of warfare. How does one measure success or failure on the battlefield? This thesis explores the effect of uncertainty on real world aerospace mission data from Operation Iraqi Freedom (OIF). Sources of uncertainty are identified during data preparation and the effects of uncertainty are investigated using multivariate analysis techniques. Two distinct cases are analyzed: one case with a low level of uncertainty in predicting whether or not a mission is pre-planned or alert generated, and another case with a high level of uncertainty in predicting whether or not a mission is a success or a failure. Three multivariate analysis techniques are used on both cases: Signal-to-Noise Ratio (SNR) saliency, Feed Forward Neural Network (FFNN) supervised training, and General Regression Neural Network (GRNN) supervised training. Measures of performance are gathered from each of these techniques and analyzed to determine the effect of uncertainty on the data.

AFIT Designator

AFIT-GOR-ENS-04-01

DTIC Accession Number

ADA422951

Download

Share

COinS