PMID: 8957688Sep 1, 1996Paper

Computational tools for the modern andrologist

Journal of Andrology
C Niederberger

Abstract

With such a wide array of computational tools to solve inference problems, andrologists and their mathematical or statistical collaborators face perhaps bewildering choices. It is tempting to criticize a method with which one is unfamiliar for its apparent complexity. Yet, many methods are quite elegant; neural computation uses nature's own best biological classifier, for example, and genetic algorithms apply rules of natural selection. Computer scientists will likely find no one single best inference engine to solve all classification problems. Rather, the modeler should choose the most appropriate computational tool based on the specific nature of a problem. If the problem can be separated into obvious components, a Markov chain may be useful. If the andrologist would like to encode a well-known clinical algorithm into the computer, the programmer may use an expert system. Once a modeler builds an inference engine, that engine is not truly useful until other andrologists use it to make inferences with their own data. Because a wide variety of computer hardware and software exists, it is a significant endeavor to translate, or "port," software designed and built on one machine to many other different computers. Fortunately, th...Continue Reading

Related Concepts

Related Feeds

Anxiety Disorders

Discover the latest research on anxiety disorders including agoraphobia, panic disorder, obsessive-compulsive disorder, and post-traumatic stress disorder here.

Related Papers

Magnetic Resonance in Medicine : Official Journal of the Society of Magnetic Resonance in Medicine
Jeffry R AlgerJohn G Curran
Telemedicine Journal : the Official Journal of the American Telemedicine Association
P B AngoodR C Merrell
Studies in Health Technology and Informatics
C CourtinP Le Beux
Acta otorrinolaringológica española
C GórrizB Arellano
© 2021 Meta ULC. All rights reserved