Skip to Main content Skip to Navigation
New interface
Conference papers

Fast classification using sparse decision DAGs

Róbert Busa-Fekete 1, 2 D. Benbouzid 3 Balázs Kégl 1, 2, 4 
3 Appstat
LAL - Laboratoire de l'Accélérateur Linéaire, LRI - Laboratoire de Recherche en Informatique
4 TAO - Machine Learning and Optimisation
LRI - Laboratoire de Recherche en Informatique, UP11 - Université Paris-Sud - Paris 11, Inria Saclay - Ile de France, CNRS - Centre National de la Recherche Scientifique : UMR8623
Abstract : In this paper we propose an algorithm that builds sparse decision DAGs (directed acyclic graphs) out of a list of base classifiers provided by an external learning method such as AdaBoost. The basic idea is to cast the DAG design task as a Markov decision process. Each instance can decide to use or to skip each base classifier, based on the current state of the classifier being built. The result is a sparse decision DAG where the base classifiers are selected in a data-dependent way. The method has a single hyperparameter with a clear semantics of controlling the accuracy/speed trade-off. The algorithm is competitive with state-of-the-art cascade detectors on three object-detection benchmarks, and it clearly outperforms them in the regime of low number of base classifiers. Unlike cascades, it is also readily applicable for multi-class classification. Using the multi-class setup, we show on a benchmark web page ranking data set that we can significantly improve the decision speed without harming the performance of the ranker.
Complete list of metadata

Cited literature [23 references]  Display  Hide  Download
Contributor : Sabine Starita Connect in order to contact the contributor
Submitted on : Friday, June 22, 2012 - 3:12:39 PM
Last modification on : Tuesday, October 25, 2022 - 4:17:05 PM
Long-term archiving on: : Sunday, September 23, 2012 - 2:41:05 AM


Files produced by the author(s)


  • HAL Id : in2p3-00711150, version 1


Róbert Busa-Fekete, D. Benbouzid, Balázs Kégl. Fast classification using sparse decision DAGs. 29th International Conference on Machine Learning (ICML 2012), Jun 2012, Edinburgh, United Kingdom. pp.951-958. ⟨in2p3-00711150⟩



Record views


Files downloads