Unsupervised Ensemble Learning with Dependent Classifiers
This addresses the challenge of combining predictions from multiple unreliable sources without labeled data, which is incremental as it builds on existing ensemble learning by accounting for dependencies.
The paper tackles the problem of unsupervised ensemble learning where classifiers may be dependent, violating the common conditional independence assumption, and introduces a statistical model and methods to detect dependencies and improve meta-learner accuracy, showing competitive performance on artificial and real datasets.
In unsupervised ensemble learning, one obtains predictions from multiple sources or classifiers, yet without knowing the reliability and expertise of each source, and with no labeled data to assess it. The task is to combine these possibly conflicting predictions into an accurate meta-learner. Most works to date assumed perfect diversity between the different sources, a property known as conditional independence. In realistic scenarios, however, this assumption is often violated, and ensemble learners based on it can be severely sub-optimal. The key challenges we address in this paper are:\ (i) how to detect, in an unsupervised manner, strong violations of conditional independence; and (ii) construct a suitable meta-learner. To this end we introduce a statistical model that allows for dependencies between classifiers. Our main contributions are the development of novel unsupervised methods to detect strongly dependent classifiers, better estimate their accuracies, and construct an improved meta-learner. Using both artificial and real datasets, we showcase the importance of taking classifier dependencies into account and the competitive performance of our approach.