Efficient Model Selection for Regularized Classification by Exploiting Unlabeled Data - Archive ouverte HAL
Communication Dans Un Congrès Année : 2015

Efficient Model Selection for Regularized Classification by Exploiting Unlabeled Data

Résumé

Hyper-parameter tuning is a resource-intensive task when optimizing classification models. The commonly used k-fold cross validation can become intractable in large scale settings when a classifier has to learn billions of parameters. At the same time, in real-world, one often encounters multi-class classification scenarios with only a few labeled examples; model selection approaches often offer little improvement in such cases and the default values of learners are used. We propose bounds for classification on accuracy and macro measures (precision, recall, F1) that motivate efficient schemes for model selection and can benefit from the existence of unlabeled data. We demonstrate the advantages of those schemes by comparing them with k-fold cross validation and hold-out estimation in the setting of large scale classification.
Fichier non déposé

Dates et versions

hal-01236587 , version 1 (01-12-2015)

Identifiants

  • HAL Id : hal-01236587 , version 1

Citer

Georgios Balikas, Ioannis Partalas, Eric Gaussier, Rohit Babbar, Massih-Reza Amini. Efficient Model Selection for Regularized Classification by Exploiting Unlabeled Data. Fourteenth International Symposium on Intelligent Data Analysis (IDA 2015), Oct 2015, Saint-Etienne, France. ⟨hal-01236587⟩
67 Consultations
0 Téléchargements

Partager

More