Our hypothesis: ensembles are well suited for problems with distribution changes.If those changes are characterizable, ensembles can be designed to tackle them.Idea: to generate different samples based on the expected distribution changes.Case study: we present ensembles versions of two binary quantification algorithms.Ensembles outperform original counterpart algorithms using trivial aggregation rules. Display Omitted Ensemble methods are widely applied to supervised learning tasks. Based on a simple strategy they often achieve good performance, especially when the single models comprising the ensemble are diverse. Diversity can be introduced into the ensemble by creating different training samples for each model. In that case, each model is trained with a data distribution that may be different from the original training set distribution. Following that idea, this paper analyzes the hypothesis that ensembles can be especially appropriate in problems that: (i) suffer from distribution changes, (ii) it is possible to characterize those changes beforehand. The idea consists in generating different training samples based on the expected distribution changes, and to train one model with each of them. As a case study, we shall focus on binary quantification problems, introducing ensembles versions for two well-known quantification algorithms. Experimental results show that these ensemble adaptations outperform the original counterpart algorithms, even when trivial aggregation rules are used.

Feeding represents the largest economic cost in meat production; therefore, selection to improve traits related to feed efficiency is a goal in most livestock breeding programs. Residual feed intake (RFI), that is, the difference between the actual and the expected feed intake based on animal's requirements, has been used as the selection criteria to improve feed efficiency since it was proposed by Kotch in 1963. In growing pigs, it is computed as the residual of the multiple regression model of daily feed intake (DFI), on average daily gain (ADG), backfat thickness (BFT), and metabolic body weight (MW). Recently, prediction using single‐output machine learning algorithms and information from SNPs as predictor variables have been proposed for genomic selection in growing pigs, but like in other species, the prediction quality achieved for RFI has been generally poor. However, it has been suggested that it could be improved through multi‐output or stacking methods. For this purpose, four strategies were implemented to predict RFI. Two of them correspond to the computation of RFI in an indirect way using the predicted values of its components obtained from (i) individual (multiple single‐output strategy) or (ii) simultaneous predictions (multi‐output strategy). The other two correspond to the direct prediction of RFI using (iii) the individual predictions of its components as predictor variables jointly with the genotype (stacking strategy), or (iv) using only the genotypes as predictors of RFI (single‐output strategy). The single‐output strategy was considered the benchmark. This research aimed to test the former three hypotheses using data recorded from 5828 growing pigs and 45,610 SNPs. For all the strategies two different learning methods were fitted: random forest (RF) and support vector regression (SVR). A nested cross‐validation (CV) with an outer 10‐folds CV and an inner threefold CV for hyperparameter tuning was implemented to test all strategies. This scheme was repeated using as predictor variables different subsets with an increasing number (from 200 to 3000) of the most informative SNPs identified with RF. Results showed that the highest prediction performance was achieved with 1000 SNPs, although the stability of feature selection was poor (0.13 points out of 1). For all SNP subsets, the benchmark showed the best prediction performance. Using the RF as a learner and the 1000 most informative SNPs as predictors, the mean (SD) of the 10 values obtained in the test sets were: 0.23 (0.04) for the Spearman correlation, 0.83 (0.04) for the zero–one loss, and 0.33 (0.03) for the rank distance loss. We conclude that the information on predicted components of RFI (DFI, ADG, MW, and BFT) does not contribute to improve the quality of the prediction of this trait in relation to the one obtained with the single‐output strategy.