Hash-Based Support Vector Machines Approximation for Large Scale Prediction
Saloua Litayem,Alexis Joly,Nozha Boujemaa +2 more
- pp 1-11
TLDR
This paper addresses the problem of speeding-up the prediction phase of linear Support Vector Machines via Locality Sensitive Hashing by building efficient hash based classifiers that are applied in a first stage in order to approximate the exact results and filter the hypothesis space.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.read more
Citations
More filters
Journal ArticleDOI
CROification: Accurate Kernel Classification with the Efficiency of Sparse Linear SVM
Mehran Kafai,Kave Eshghi +1 more
TL;DR: A new kernel is introduced, the CRO (Concomitant Rank Order) kernel that approximates the Gaussian kernel on the unit sphere and a randomized feature map is introduced that produces sparse, high-dimensional feature vectors whose inner product asymptotically equals theCRO kernel.
Journal ArticleDOI
Towards large-scale multimedia retrieval enriched by knowledge about human interpretation
TL;DR: This paper defends the importance of human-machine cooperation which incorporates the above knowledge into LSMR, and defines its three future directions (cognition-based, ontology-based and adaptive learning) depending on types of knowledge, and suggest to explore each direction by considering its relation to the others.
Journal ArticleDOI
Scalable Mobile Visual Classification by Kernel Preserving Projection Over High-Dimensional Features
TL;DR: This work proposes an unsupervised linear dimension reduction algorithm, kernel preserving projection (KPP), which approximates the kernel matrix of high dimensional features with low dimensional linear embedding and proves that the proposed method outperforms existing dimension reduction methods.
Book ChapterDOI
Large-Scale R-CNN with Classifier Adaptive Quantization
Ryota Hinami,Shin'ichi Satoh +1 more
TL;DR: This paper presents a novel quantization method designed forlinear classification wherein the quantization error is re-defined for linear classification and approximates the error as the empirical error with pre-defined multiple exemplar classifiers and captures the variance and common attributes of object category classifiers effectively.
Journal ArticleDOI
A Systematic Review on Minwise Hashing Algorithms
Jingjing Tang,Yingjie Tian +1 more
TL;DR: The purpose of this paper is to review minwise hashing algorithms in detail and provide an insightful understanding of current developments and their limitations, major opportunities and challenges, extensions and variants as well as potential important research directions have been pointed out.
References
More filters
Proceedings ArticleDOI
ImageNet: A large-scale hierarchical image database
TL;DR: A new database called “ImageNet” is introduced, a large-scale ontology of images built upon the backbone of the WordNet structure, much larger in scale and diversity and much more accurate than the current image datasets.
Journal ArticleDOI
Bagging predictors
TL;DR: Tests on real and simulated data sets using classification and regression trees and subset selection in linear regression show that bagging can give substantial gains in accuracy.
Journal Article
LIBLINEAR: A Library for Large Linear Classification
TL;DR: LIBLINEAR is an open source library for large-scale linear classification that supports logistic regression and linear support vector machines and provides easy-to-use command-line tools and library calls for users and developers.
Proceedings ArticleDOI
Similarity estimation techniques from rounding algorithms
TL;DR: It is shown that rounding algorithms for LPs and SDPs used in the context of approximation algorithms can be viewed as locality sensitive hashing schemes for several interesting collections of objects.
Proceedings ArticleDOI
Training linear SVMs in linear time
TL;DR: A Cutting Plane Algorithm for training linear SVMs that provably has training time 0(s,n) for classification problems and o(sn log (n)) for ordinal regression problems and several orders of magnitude faster than decomposition methods like svm light for large datasets.