Ensemble methods like bagging and boosting that combine the decisions of multiple hypotheses are some of the strongest existing machine learning methods. The diversity of the memb...
Learning Classifier Systems use evolutionary algorithms to facilitate rule- discovery, where rule fitness is traditionally payoff based and assigned under a sharing scheme. Most c...
In this paper we apply a machine learning approach to the problem of estimating the number of defects called Regression via Classification (RvC). RvC initially automatically discr...
Stamatia Bibi, Grigorios Tsoumakas, Ioannis Stamel...
In this paper we present lessons learned in the Evaluating Predictive Uncertainty Challenge. We describe the methods we used in regression challenges, including our winning method ...
We studied 1372 LacI-family transcription factors and their 4484 DNA binding sites using machine learning algorithms and feature selection techniques. The Naive Bayes classifier a...