“…Popular ensemble approaches include several variants of bagging and boosting algorithms, which have been shown to provide very good results in several cases (Abell an & Castellano, 2017; Bequ e & Lessmann, 2017; Finlay, 2011; Marqu es, Garc ıa, & S anchez, 2012). Hybrid systems, which rely on the combination of different techniques for feature/sample selection and model fitting as well as different modeling schemes (Doumpos, Niklis, Zopounidis, & Andriosopoulos, 2015;Niklis, Doumpos, & Zopounidis, 2014;Oreski, Oreski, & Oreski, 2012;Yeh, Lin, & Hsu, 2012;Yu, Wang, & Lai, 2009;Zhang, Gao, & Shi, 2014).…”