Measuring the VC-dimension of a learning machine
Neural Computation
The nature of statistical learning theory
The nature of statistical learning theory
Machine Learning
From noise-free to noise-tolerant and from on-line to batch learning
COLT '95 Proceedings of the eighth annual conference on Computational learning theory
Large margin classification using the perceptron algorithm
COLT' 98 Proceedings of the eleventh annual conference on Computational learning theory
Generalization performance of support vector machines and other pattern classifiers
Advances in kernel methods
Boosting the margin: A new explanation for the effectiveness of voting methods
ICML '97 Proceedings of the Fourteenth International Conference on Machine Learning
Bayesian Classifiers Are Large Margin Hyperplanes in a Hilbert Space
ICML '98 Proceedings of the Fifteenth International Conference on Machine Learning
A Note on a Scale-Sensitive Dimension of Linear Bounded Functionals in Banach Spaces
ALT '97 Proceedings of the 8th International Conference on Algorithmic Learning Theory
Generalization Performance of Classifiers in Terms of Observed Covering Numbers
EuroCOLT '99 Proceedings of the 4th European Conference on Computational Learning Theory
Estimation of Dependences Based on Empirical Data: Springer Series in Statistics (Springer Series in Statistics)
Structural risk minimization over data-dependent hierarchies
IEEE Transactions on Information Theory
Support vector machines: hype or hallelujah?
ACM SIGKDD Explorations Newsletter - Special issue on “Scalable data mining algorithms”
Tree Decomposition for Large-Scale SVM Problems
The Journal of Machine Learning Research
| Hi-index | 0.00 |
A number of results have bounded generalization of a classi fier in terms of its margin on the training points. There has been some debate about whether the minimum margin is the best measure of the distribution of training set margin values with which to estimate the generalization. Freund and Schapire [6] have shown how a different function of the margin distribution can be used to bound the number of mistakes of an on-line learning algorithm for a perceptron, as well as an expected error bound. We show that a slight generalization of their construction can be used to give a pac style bound on the tail of the distribution of the generalization errors that arise from a given sample size. We also derive an algorithm for optimizing the new measure for general kernel based learning machines. Some preliminary experiments are presented.