Discovering Risk of Disease with a Learning Classifier System

A learning classifier system, EpiCS, was used to de rive a continuous measure of disease risk in a series of 250 individuals. Using the area under the receiver-ope rating characteristic curve, this measure was compared wit h the risk estimate derived for the same individuals by logistic regression. Over 20 training-testing trial s, risk estimates derived by EpiCS were consistently more accurate (mean area=0.97, SD=0.01) than that derive d by logistic regression (mean area=0.89, SD=0.02). The areas for the trials with minimum and maximum classification performance on testing were signific antly greater (p=0.019 and p<0.001, respectively) than th e area for the logistic regression curve. This inves tigation demonstrated the ability of a learning classifier s ystem to produce output that is clinically meaningful in diagnostic classification.

[1]  J. Goldkrand,et al.  Prediction of respiratory distress syndrome by a new colorimetric assay. , 1992, American journal of obstetrics and gynecology.

[2]  M. Graffar.  [Modern epidemiology]. , 1971, Bruxelles medical.

[3]  L L Coultrip,et al.  Evaluation of rapid diagnostic tests in the detection of microbial invasion of the amniotic cavity. , 1992, American journal of obstetrics and gynecology.

[4]  N. Whitworth,et al.  Receiver-operator characteristic, efficiency analysis, and predictive value of serum progesterone concentration as a test for abnormal gestations. , 1992, American journal of obstetrics and gynecology.

[5]  F Van Lente,et al.  The predictive value of serum enzymes for perioperative myocardial infarction after cardiac operations. An autopsy study. , 1989, The Journal of thoracic and cardiovascular surgery.

[6]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[7]  E Somoza,et al.  Evaluation and optimization of diagnostic tests using receiver operating characteristic analysis and information theory. , 1989, International journal of bio-medical computing.

[8]  J. Hanley,et al.  The meaning and use of the area under a receiver operating characteristic (ROC) curve. , 1982, Radiology.

[9]  Sandip Sen,et al.  Newboole: A Fast GBML System , 1990, ML.

[10]  Machine Learning, Proceedings of the Seventh International Conference on Machine Learning, Austin, Texas, USA, June 21-23, 1990 , 1990, ICML.

[11]  David Hume,et al.  A treatise of human nature (1739-40) , 1996 .

[12]  D. M. Green,et al.  Signal detection theory and psychophysics , 1966 .

[13]  J. Hanley,et al.  Statistical Approaches to the Analysis of Receiver Operating Characteristic (ROC) Curves , 1984, Medical decision making : an international journal of the Society for Medical Decision Making.

[14]  D Gur,et al.  Comparing imaging systems by ROC studies. Detection versus interpretation. , 1989, Investigative radiology.

[15]  O Pahlm,et al.  Decision rules for the ECG diagnosis of inferior myocardial infarction. , 1990, Computers and biomedical research, an international journal.