TY  - CONF
ID  - ln:MLCaggregation
T1  - On Aggregation in Ensembles of Multilabel Classifiers
A1  - Nguyen, Vu-Linh
A1  - Hüllermeier, Eyke
A1  - Rapp, Michael
A1  - Loza Mencía, Eneldo
A1  - Fürnkranz, Johannes
ED  - Appice, Annalisa
ED  - Tsoumakas, Grigorios
ED  - Manolopoulos, Yannis
ED  - Matwin, Stan
TI  - Discovery Science
T3  - Lecture Notes in Computer Science
Y1  - 2020
VL  - 12323
SP  - 533
EP  - 547
PB  - Springer International Publishing
AD  - Cham
SN  - 978-3-030-61527-7
UR  - https://arxiv.org/abs/2006.11916
M2  - doi: https://doi.org/10.1007/978-3-030-61527-7_35
KW  - Combine then Predict
KW  - Ensembles of Multilabel Classifiers
KW  - F-measure
KW  - Hamming loss
KW  - Predict then Combine
KW  - Subset 0/1 loss
N2  - While a variety of ensemble methods for multilabel classification have been proposed in the literature, the question of how to aggregate the predictions of the individual members of the ensemble has received little attention so far. In this paper, we introduce a formal framework of ensemble multilabel classification, in which we distinguish two principal approaches: "predict then combine" (PTC), where the ensemble members first make loss minimizing predictions which are subsequently combined, and "combine then predict" (CTP), which first aggregates information such as marginal label probabilities from the individual ensemble members, and then derives a prediction from this aggregation. While both approaches generalize voting techniques commonly used for multilabel ensembles, they allow to explicitly take the target performance measure into account. Therefore, concrete instantiations of CTP and PTC can be tailored to concrete loss functions. Experimentally, we show that standard voting techniques are indeed outperformed by suitable instantiations of CTP and PTC, and provide some evidence that CTP performs well for decomposable loss functions, whereas PTC is the better choice for non-decomposable losses.
ER  -