ID  - jf:TUD-KE-2007-05
T1  - Pairwise Learning of Multilabel Classifications with Perceptrons
A1  - Loza Mencía, Eneldo
A1  - Fürnkranz, Johannes
Y1  - 2007
IS  - TUD-KE-2007-05
T2  - TU Darmstadt, Knowledge Engineering Group
UR  - http://www.ke.tu-darmstadt.de/publications/reports/tud-ke-2007-05.pdf
N2  - Multiclass Multilabel Perceptrons (MMP) are an efficient incremental algorithm for
training a team of perceptrons for a multilabel prediction task. The key idea is to train
one binary classifier per label, as is typically done for addressing multilabel problems, but
to make the training signal dependent on the performance of the whole ensemble. In this
paper, we propose an alternative approach that is based on a pairwise approach, i.e., we
incrementally train a perceptron for each pair of classes. An evaluation on the Reuters
2000 (RCV1) data shows that our multilabel pairwise perceptron (MLPP) algorithm yields
substantial improvements over MMP in terms of ranking quality and overfitting resistance,
while maintaining its efficiency. Despite the quadratic increase in the number of perceptrons
that have to be trained, the increase in computational complexity is bounded by the average
number of labels per training example.
ER  -