Submitting Campus

Daytona Beach


Physical Sciences

Document Type


Publication/Presentation Date



We briefly review the work of the past decade on automated classification of stellar spectra and discuss techniques which show par­ticular promise. Emphasis is placed on Artificial Neural Network and Principle Component Analysis based techniques, due both to our greater familiarity with these and to their rising popularity. As an example of the abilities of current techniques we report on our automated classification work based on the visual classifications of N. Houk (Michigan Spectral Catalogue, Vol. 1 - 4, 1975, 1978, 1982, 1988).

Publication Title

ASP Conference Series


Astronomical Society of the Pacific

Additional Information

Dr. von Hippel was not affiliated with Embry-Riddle Aeronautical University at the time this paper was published.