Relaxation of hard classification targets for LSE minimization

Authors:
Kar-Ann Toh;Xudong Jiang;Wei-Yun Yau
Affiliations:
Biometrics Engineering Research Center (BERC), School of Electrical & Electronic Engineering, Yonsei University, Seoul, Korea;School of Electrical & Electronic Engineering, Nanyang Technological University, Singapore;Institute for Infocomm Research, Singapore
Venue:
EMMCVPR'05 Proceedings of the 5th international conference on Energy Minimization Methods in Computer Vision and Pattern Recognition
Year:
2005

Citing 9
Cited 0

The nature of statistical learning theory

The nature of statistical learning theory
Pattern classification: a unified view of statistical and neural approaches

Pattern classification: a unified view of statistical and neural approaches
Partial Classification: The Benefit of Deferred Decision

IEEE Transactions on Pattern Analysis and Machine Intelligence
Adaptive Sparseness for Supervised Learning

IEEE Transactions on Pattern Analysis and Machine Intelligence
Sparse bayesian learning and the relevance vector machine

The Journal of Machine Learning Research
Pattern Classification (2nd Edition)

Pattern Classification (2nd Edition)
A Meta-Learning Method to Select the Kernel Width in Support Vector Regression

Machine Learning
Benchmarking a Reduced Multivariate Polynomial Pattern Classifier

IEEE Transactions on Pattern Analysis and Machine Intelligence
A reduced multivariate polynomial model for multimodal biometrics and classifiers fusion

IEEE Transactions on Circuits and Systems for Video Technology

Quantified Score

Hi-index	0.00

Visualization

Abstract

In the spirit of stabilizing a solution to handle possible over-fitting of data which is especially common for high order models, we propose a relaxed target training method for regression models which are linear in parameters. This relaxation of training target from the conventional binary values to disjoint classification spaces provides good classification fidelity according to a threshold treatment during the decision process. A particular design to relax the training target is provided under practical consideration. Extension to multiple class problems is formulated before the method is applied to a plug-in full multivariate polynomial model and a reduced model on synthetic data sets to illustrate the idea. Additional experiments were performed using real-world data from the UCI[1] data repository to derive certain empirical evidence.