Hash-Based Support Vector Machines Approximation for Large Scale Prediction

Saloua Litayem Ouertani 1, 2 Alexis Joly 2 Nozha Boujemaa 3
2 ZENITH - Scientific Data Management
LIRMM - Laboratoire d'Informatique de Robotique et de Microélectronique de Montpellier, CRISAM - Inria Sophia Antipolis - Méditerranée
Abstract : How-to train effective classifiers on huge amount of multimedia data is clearly a major challenge that is attracting more and more research works across several communities. Less efforts however are spent on the counterpart scalability issue: how to apply big trained models efficiently on huge non annotated media collections ? In this paper, we address the problem of speeding-up the prediction phase of linear Support Vector Machines via Locality Sensitive Hashing. We propose building efficient hash based classifiers that are applied in a first stage in order to approximate the exact results and filter the hypothesis space. Experiments performed with millions of one-against-one classifiers show that the proposed hash-based classifier can be more than two orders of magnitude faster than the exact classifier with minor losses in quality.
Keywords : hashing methods LSH SVM
Complete list of metadatas

Cited literature [20 references]  Display  Hide  Download

https://hal.inria.fr/hal-00733912
Contributor : Saloua Litayem Ouertani <>
Submitted on : Thursday, September 20, 2012 - 10:08:46 AM
Last modification on : Thursday, February 14, 2019 - 11:06:11 AM
Long-term archiving on: Friday, December 21, 2012 - 3:49:28 AM

File

bmvc_final.pdf
Publisher files allowed on an open archive

Identifiers

  • HAL Id : hal-00733912, version 1

Collections

Citation

Saloua Litayem Ouertani, Alexis Joly, Nozha Boujemaa. Hash-Based Support Vector Machines Approximation for Large Scale Prediction. BMVC: British Machine Vision Conference, Sep 2012, Surrey, United Kingdom. pp.86.1-86.11. ⟨hal-00733912⟩

Share

Metrics

Record views

1028

Files downloads

554