Using domain-specific knowledge in generalization error bounds for support vector machine learning

Authors:
Enes Eryarsoy;Gary J. Koehler;Haldun Aytug
Affiliations:
Faculty of Management, Sabanci University, of Management, Sabanci University, Orhanli, Istanbul, 34956, Turkey;Department of Information Systems and Operations Management, Warrington College of Business, University of Florida, 351 STZ, P.O. Box 117169, Gainesville, Florida 32611-7169, USA;Department of Information Systems and Operations Management, Warrington College of Business, University of Florida, 351 STZ, P.O. Box 117169, Gainesville, Florida 32611-7169, USA
Venue:
Decision Support Systems
Year:
2009

Citing 20
Cited 2

A theory of the learnable

Communications of the ACM
Scale-sensitive dimensions, uniform convergence, and learnability

Journal of the ACM (JACM)
Prior knowledge in support vector kernels

NIPS '97 Proceedings of the 1997 conference on Advances in neural information processing systems 10
On domain knowledge and feature selection using a support vector machine

Pattern Recognition Letters
Estimating the number of vertices of a polyhedron

Information Processing Letters
An introduction to support Vector Machines: and other kernel-based learning methods

An introduction to support Vector Machines: and other kernel-based learning methods
Support vector machines: hype or hallelujah?

ACM SIGKDD Explorations Newsletter - Special issue on “Scalable data mining algorithms”
Obstacle Collision Detection Using Best Ellipsoid Fit

Journal of Intelligent and Robotic Systems
Latent Semantic Kernels

Journal of Intelligent Information Systems
Text Categorization with Suport Vector Machines: Learning with Many Relevant Features

ECML '98 Proceedings of the 10th European Conference on Machine Learning
Composite Kernels for Hypertext Categorisation

ICML '01 Proceedings of the Eighteenth International Conference on Machine Learning
Credit rating analysis with support vector machines and neural networks: a market comparative study

Decision Support Systems - Special issue: Data mining for financial decision making
Estimation of Dependences Based on Empirical Data: Springer Series in Statistics (Springer Series in Statistics)

Estimation of Dependences Based on Empirical Data: Springer Series in Statistics (Springer Series in Statistics)
Simpler knowledge-based support vector machines

ICML '06 Proceedings of the 23rd international conference on Machine learning
A process model to develop an internal rating system: sovereign credit ratings

Decision Support Systems
An intelligent information agent for document title classification and filtering in document-intensive domains

Decision Support Systems
A machine learning approach to web page filtering using content and structure analysis

Decision Support Systems
The lack of a priori distinctions between learning algorithms

Neural Computation
The existence of a priori distinctions between learning algorithms

Neural Computation
Predicting going concern opinion with data mining

Decision Support Systems

Machine learning and genetic algorithms in pharmaceutical development and manufacturing processes

Decision Support Systems
Using structure-based data transformation method to improve prediction accuracies for small data sets

Decision Support Systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this study we describe a methodology to exploit a specific type of domain knowledge in order to find tighter error bounds on the performance of classification via Support Vector Machines. The domain knowledge we consider is that the input space lies inside of a specified convex polytope. First, we consider prior knowledge about the domain by incorporating upper and lower bounds of attributes. We then consider a more general framework that allows us to encode prior knowledge in the form of linear constraints formed by attributes. By using the ellipsoid method from optimization literature, we show that, this can be exploited to upper bound the radius of the hyper-sphere that contains the input space, and enables us to tighten generalization error bounds. We provide a comparative numerical analysis and show the effectiveness of our approach.