CLINICAL LABORATORY TEST EVALUATION USING FEATURE SELECTION TECHNIQUES.

John W. Lewis, James C. Boyd, Henry J. Votava, Jack H. Ladenson

Research output: Contribution to conferencePaperpeer-review

2 Scopus citations

Abstract

Three feature selection techniques were used to estimate the relative utility of each of 15 laboratory tests on a data base of 316 patients with disorders of calcium homeostasis. Comparison of the relative utilities given by the feature selection methods indicated differences among the techniques as well as some findings not predicted by current medical thought. Although the preliminary results reported here are promising, further studies are needed using additional feature selection methodologies.

Original languageEnglish
Pages47-54
Number of pages8
StatePublished - Jan 1 1977
EventIEEE Comput Soc Conf on Pattern Recognition and Image Process, Proc - Troy, NY, USA
Duration: Jun 6 1977Jun 8 1977

Conference

ConferenceIEEE Comput Soc Conf on Pattern Recognition and Image Process, Proc
CityTroy, NY, USA
Period06/6/7706/8/77

Fingerprint

Dive into the research topics of 'CLINICAL LABORATORY TEST EVALUATION USING FEATURE SELECTION TECHNIQUES.'. Together they form a unique fingerprint.

Cite this