Date of Award
6-1990
Degree Name
Master of Science
Department
Computer Science
First Advisor
Dr. Robert Trenary
Second Advisor
Dr. Ben Pinkowski
Third Advisor
Dr. Donna Kaminski
Access Setting
Masters Thesis-Open Access
Abstract
This thesis traces the development of a neural net model which will solve the assignment problem. In addition, the neural net model was compared for optimal performance with four other algorithms: two greedy methods and two well-known solutions, Vogel's Approximation Method (VAM) and the Hungarian method. Assignments were made to randomly generated problems using each method and the results were compared.
The findings from this testing indicated that: (a) the assignments selected by the neural net model were closer to optimal than the two greedy methods and VAM, and (b) the assignments selected by the neural net were not optimal. It was concluded that the neural net model, although not optimal, could have useful applications.
Recommended Citation
Moss, Philip Paul, "Development of a Neural Net Model for the Assignment Problem" (1990). Masters Theses. 1090.
https://scholarworks.wmich.edu/masters_theses/1090