Academic
Publications
Reliable Confidence Measures for Medical Diagnosis With Evolutionary Algorithms

Reliable Confidence Measures for Medical Diagnosis With Evolutionary Algorithms,10.1109/TITB.2010.2091144,IEEE Transactions on Information Technology

Reliable Confidence Measures for Medical Diagnosis With Evolutionary Algorithms   (Citations: 1)
BibTex | RIS | RefWorks Download
Conformal Predictors (CPs) are machine learning al- gorithms that can provide predictions complemented with valid confidence measures. In medical diagnosis, such measures are highly desirable, as medical experts can gain additional informa- tion for each machine diagnosis. A risk assessment in each predic- tion can play an important role for medical decision making, in which the outcome can be critical for the patients. Several classical machine learning methods can be incorporated into the CP frame- work. In this paper, we propose a CP that makes use of evolved rule sets generated by a genetic algorithm (GA). The rule-based GA has the advantage of being human readable. We apply our method on two real-world datasets for medical diagnosis, one dataset for breast cancer diagnosis, which contains data gathered from fine needle aspirate of breast mass; and one dataset for ovarian cancer diagnosis, which contains proteomic patterns identified in serum. Our results on both datasets show that the proposed method is as accurate as the classical techniques, while it provides reliable and useful confidence measures. Index Terms—Breast cancer, confidence values, conformal pre- diction (CP), genetic algorithms (GAs), medical diagnosis, ovarian cancer.
Journal: IEEE Transactions on Information Technology in Biomedicine - TITB , vol. 15, no. 1, pp. 93-99, 2011
Cumulative Annual
View Publication
The following links allow you to view full publications. These links are maintained by other sources not affiliated with Microsoft Academic Search.
Sort by: