{"publisher":"Springer","oa":1,"scopus_import":1,"publication_status":"published","month":"10","issue":"1-2","title":"Learning a priori constrained weighted majority votes","_id":"2180","doi":"10.1007/s10994-014-5462-z","type":"journal_article","citation":{"chicago":"Bellet, Aurélien, Amaury Habrard, Emilie Morvant, and Marc Sebban. “Learning a Priori Constrained Weighted Majority Votes.” <i>Machine Learning</i>. Springer, 2014. <a href=\"https://doi.org/10.1007/s10994-014-5462-z\">https://doi.org/10.1007/s10994-014-5462-z</a>.","mla":"Bellet, Aurélien, et al. “Learning a Priori Constrained Weighted Majority Votes.” <i>Machine Learning</i>, vol. 97, no. 1–2, Springer, 2014, pp. 129–54, doi:<a href=\"https://doi.org/10.1007/s10994-014-5462-z\">10.1007/s10994-014-5462-z</a>.","ieee":"A. Bellet, A. Habrard, E. Morvant, and M. Sebban, “Learning a priori constrained weighted majority votes,” <i>Machine Learning</i>, vol. 97, no. 1–2. Springer, pp. 129–154, 2014.","ista":"Bellet A, Habrard A, Morvant E, Sebban M. 2014. Learning a priori constrained weighted majority votes. Machine Learning. 97(1–2), 129–154.","ama":"Bellet A, Habrard A, Morvant E, Sebban M. Learning a priori constrained weighted majority votes. <i>Machine Learning</i>. 2014;97(1-2):129-154. doi:<a href=\"https://doi.org/10.1007/s10994-014-5462-z\">10.1007/s10994-014-5462-z</a>","apa":"Bellet, A., Habrard, A., Morvant, E., &#38; Sebban, M. (2014). Learning a priori constrained weighted majority votes. <i>Machine Learning</i>. Springer. <a href=\"https://doi.org/10.1007/s10994-014-5462-z\">https://doi.org/10.1007/s10994-014-5462-z</a>","short":"A. Bellet, A. Habrard, E. Morvant, M. Sebban, Machine Learning 97 (2014) 129–154."},"oa_version":"Submitted Version","date_created":"2018-12-11T11:56:10Z","publication":"Machine Learning","date_published":"2014-10-01T00:00:00Z","year":"2014","author":[{"full_name":"Bellet, Aurélien","first_name":"Aurélien","last_name":"Bellet"},{"last_name":"Habrard","first_name":"Amaury","full_name":"Habrard, Amaury"},{"last_name":"Morvant","id":"4BAC2A72-F248-11E8-B48F-1D18A9856A87","full_name":"Morvant, Emilie","orcid":"0000-0002-8301-7240","first_name":"Emilie"},{"last_name":"Sebban","full_name":"Sebban, Marc","first_name":"Marc"}],"publist_id":"4802","language":[{"iso":"eng"}],"department":[{"_id":"ChLa"}],"page":"129 - 154","user_id":"4435EBFC-F248-11E8-B48F-1D18A9856A87","project":[{"call_identifier":"FP7","grant_number":"308036","_id":"2532554C-B435-11E9-9278-68D0E5697425","name":"Lifelong Learning of Visual Scene Understanding"}],"volume":97,"status":"public","abstract":[{"lang":"eng","text":"Weighted majority votes allow one to combine the output of several classifiers or voters. MinCq is a recent algorithm for optimizing the weight of each voter based on the minimization of a theoretical bound over the risk of the vote with elegant PAC-Bayesian generalization guarantees. However, while it has demonstrated good performance when combining weak classifiers, MinCq cannot make use of the useful a priori knowledge that one may have when using a mixture of weak and strong voters. In this paper, we propose P-MinCq, an extension of MinCq that can incorporate such knowledge in the form of a  constraint over the distribution of the weights, along with general proofs of convergence that stand in the sample compression setting for data-dependent voters. The approach is applied to a vote of k-NN classifiers with a specific modeling of the voters' performance. P-MinCq significantly outperforms the classic k-NN classifier, a symmetric NN and MinCq using the same voters. We show that it is also competitive with LMNN, a popular metric learning algorithm, and that combining both approaches further reduces the error."}],"intvolume":"        97","date_updated":"2021-01-12T06:55:49Z","quality_controlled":"1","day":"01","acknowledgement":"This work was funded by the French project SoLSTiCe ANR-13-BS02-01 of the ANR. ","ec_funded":1,"main_file_link":[{"open_access":"1","url":"https://hal.archives-ouvertes.fr/hal-01009578/document"}]}