Speeding up the solution of multilabel problems with Support Vector Machines
Lists
Only you see this
Abstract
The classical SVM approach to solve multilabel\nproblems consists in training a single classifier for each class.\nWe propose a compact model that considers the whole set of\nclassifiers at once. Our strategy focuses on the shared use of\nthe kernel matrix information between different classifiers in\norder to reduce the complexity of the learning task. Experiments\nwith the Reuters-21578 corpus show a speedup in term of\nkernel computations (cache misses) preserving state-of-the-art\nperformance.
BibTeX
@inproceedings{aiolli2003speeding,
author = {Fabio Aiolli and Filippo Portera and Alessandro Sperduti},
title = {{Speeding up the solution of multilabel problems with Support Vector Machines}},
booktitle = {Research},
year = {2003},
}Loading related content...