Premium
Experiments in machine learning using artificial neural networks for control and image compression
Author(s) -
Watta Paul B.,
Hassoun Mohamad H.
Publication year - 1995
Publication title -
computer applications in engineering education
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.478
H-Index - 29
eISSN - 1099-0542
pISSN - 1061-3773
DOI - 10.1002/cae.6180030307
Subject(s) - computer science , artificial neural network , bottle neck , image compression , artificial intelligence , control (management) , task (project management) , process (computing) , machine learning , image processing , image (mathematics) , bottle , engineering , mechanical engineering , systems engineering , operating system
This article presents two computer projects which can be incorporated in an undergraduate course in artificial neural networks. These projects allow students to explore both the exciting possibilities and practical limitations of machine learning using the popular back‐prop algorithm. In the first project, a neural net is trained in real‐time to control a ball balanced on a beam by observing the actions of a human operator. In the second project, a bottle‐neck—type neural net is trained for the task of image compression. Once completed, both these projects can be used to visually demonstrate both the process and power of machine learning to a nontechnical audience, including, for example, non‐engineering undergraduate students, and even high‐school students.