Simpler knowledge-based support vector machines

Authors:
Quoc V. Le;Alex J. Smola;Thomas Gärtner
Affiliations:
Australian National University, ACT, Australia and Statistical Machine Learning Program, National ICT Australia, Australia;Australian National University, ACT, Australia and Statistical Machine Learning Program, National ICT Australia, Australia;Fraunhofer AIS.KD, Schloß Birlinghoven, Germany
Venue:
ICML '06 Proceedings of the 23rd international conference on Machine learning
Year:
2006

Citing 9
Cited 8

The connection between regularization operators and support vector kernels

Neural Networks
Bayesian Learning for Neural Networks

Bayesian Learning for Neural Networks
The concave-convex procedure

Neural Computation
Algorithmic luckiness

The Journal of Machine Learning Research
Minimal kernel classifiers

The Journal of Machine Learning Research
Rademacher and gaussian complexities: risk bounds and structural results

The Journal of Machine Learning Research
Information Theory, Inference & Learning Algorithms

Information Theory, Inference & Learning Algorithms
Knowledge-Based Kernel Approximation

The Journal of Machine Learning Research
New Support Vector Algorithms

Neural Computation

Nonparametric Quantile Estimation

The Journal of Machine Learning Research
Mutually beneficial learning with application to on-line news classification

Proceedings of the ACM first Ph.D. workshop in CIKM
Using domain-specific knowledge in generalization error bounds for support vector machine learning

Decision Support Systems
gBoost: a mathematical programming approach to graph classification and regression

Machine Learning
Refining rules incorporated into knowledge-based support vector learners via successive linear programming

AAAI'07 Proceedings of the 22nd national conference on Artificial intelligence - Volume 1
Incorporating a priori knowledge from detractor points into support vector classification

ICANNGA'11 Proceedings of the 10th international conference on Adaptive and natural computing algorithms - Volume Part II
Learning with box kernels

ICONIP'11 Proceedings of the 18th international conference on Neural Information Processing - Volume Part II
A theoretical framework for supervised learning from regions

Neurocomputing

Quantified Score

Hi-index	0.00

Visualization

Abstract

If appropriately used, prior knowledge can significantly improve the predictive accuracy of learning algorithms or reduce the amount of training data needed. In this paper we introduce a simple method to incorporate prior knowledge in support vector machines by modifying the hypothesis space rather than the optimization problem. The optimization problem is amenable to solution by the constrained concave convex procedure, which finds a local optimum. The paper discusses different kinds of prior knowledge and demonstrates the applicability of the approach in some characteristic experiments.