Adaptivity of support vector machines with l1 penalty

Publication Year: 2004

Type

Unknown Type

Authors

Advanced Analysis ...

Results

Collaboration

Competence

Consortium

Com. Trends

Coll. Trends

Tarigan Bernadetta

van de Geer Sara

Abstract

We consider the problem of adaptation to model complexity and noise level, when using support vector machine loss in binary classification. We show that with $\ell_1$ complexity regularization, adaptive rates can be obtained.

Project

Results

PASCAL Project

Organisations

Results

Leiden University

Related Publications (similar authors)

Results

0.500 Square root penalty: adaptation to the margin in classification and in edge estimation

Related Publications (similar projects)

Advanced Analysis ...

Results

Collaboration

Competence

Com. Trends

1.000 Inlier-Based ICA with an Application to Super-imposed Images

1.000 LS-SVM hyperparameter selection with a nonparametric noise estimator

1.000 Growing an n-gram model

1.000 Using KCCA for Japanese-English cross-language information retrieval and classification

1.000 Combining a Filter Method with SVMs

1.000 Unsupervised Learning with Term Clustering for Thematic segmentation of Texts .

1.000 Kernel Based Depdendence Detection in the Macaque Visual Cortex

1.000 Feature Selection Methods involving Support Vector Machines for Prediction of Insolvency in Non-life Insurance Companies

1.000 Protein ranking: from local to global structure in the protein similarity network

1.000 Kernel Constrained Covariance for Dependence Measurement

1 2 3 4 5 6 7 8 9 10 ...

Related Publications (similar organisations)

Results

1.000 Square root penalty: adaptation to the margin in classification and in edge estimation

Datasets

Pascal