Learning optimal classification trees using a binary linear program formulation (extended abstract)

Sicco Verwer, Yingqian Zhang

Onderzoeksoutput: Bijdrage aan congresAbstractAcademic

3 Downloads (Pure)

Samenvatting

We provide a new formulation for the problem of learning the optimal classification tree of a given depth as a binary linear program. A limitation of previously proposed Mathematical Optimization formulations is that they create constraints and variables for every row in the training data. As a result, the running time of the existing Integer Linear programming (ILP) formulations increases dramatically with the size of data. In our new binary formulation, we aim to circumvent this problem by making the formulation size largely independent from the training data size. We show experimentally that our formulation achieves better performance than existing formulations on both small and large problem instances within shorter running time.
Originele taal-2Engels
Aantal pagina's2
StatusGepubliceerd - 1 jan 2019
Evenement31st Benelux Conference on Artificial Intelligence and the 28th Belgian Dutch Conference on Machine Learning, BNAIC/BENELEARN 2019 - Brussels, België
Duur: 6 nov 20198 nov 2019

Congres

Congres31st Benelux Conference on Artificial Intelligence and the 28th Belgian Dutch Conference on Machine Learning, BNAIC/BENELEARN 2019
LandBelgië
StadBrussels
Periode6/11/198/11/19

Vingerafdruk Duik in de onderzoeksthema's van 'Learning optimal classification trees using a binary linear program formulation (extended abstract)'. Samen vormen ze een unieke vingerafdruk.

  • Citeer dit

    Verwer, S., & Zhang, Y. (2019). Learning optimal classification trees using a binary linear program formulation (extended abstract). Abstract van 31st Benelux Conference on Artificial Intelligence and the 28th Belgian Dutch Conference on Machine Learning, BNAIC/BENELEARN 2019, Brussels, België.