Ilmari Kylliäinen
Department of Digital Humanities, University of Helsinki, Finland
Miikka Silfverberg
Department of Digital Humanities, University of Helsinki, Finland
Download article
Published in: Proceedings of the 22nd Nordic Conference on Computational Linguistics (NoDaLiDa), September 30 - October 2, Turku, Finland
Linköping Electronic Conference Proceedings 167:32, p. 304--309
NEALT Proceedings Series 42:32, p. 304--309
Published: 2019-10-02
ISBN: 978-91-7929-995-8
ISSN: 1650-3686 (print), 1650-3740 (online)
We investigate different ensemble learning techniques for neural morphological inflection using bidirectional LSTM encoder-decoder models with attention. We experiment with weighted and unweighted majority voting and bagging. We find that all investigated ensemble methods lead to improved accuracy over a baseline of a single model. However, contrary to expectation based on earlier work by \newcite{Najafi2018} and \newcite{Silfverberg2017}, weighting does not deliver clear benefits. Bagging was found to underperform plain voting ensembles in general.
morphological inflection
computational morphology
ensemble learning
deep learning
recurrent neural networks
