Meta learning on small biomedical datasets

Abstract

Meta-learning is one of subsections of supervised machine learning that has continuously grown with interests to apply on new data sets in the late years. Meta learning is the process of knowledge that is acquired by the examples. Bagging, dagging, decorate, rotation forest, and filtered classifiers are well known meta-learning algorithms that are performed to compare with these meta-learning algorithms on 8 different biomedical datasets. In these algorithms, the rotation forest had the better results according to F-measurement and ROC Area in most cases.