THE RELATIONSHIP BETWEEN KERNEL AND CLASSIFIER FUSION IN KERNEL-BASED MULTI-MODAL PATTERN RECOGNITION: AN EXPERIMENTAL STUDY

摘要：

Two distinct principles of multi-modal kernel-based pattern recognition, kernel and classifier fusion, are demonstrated to share common underlying characteristics via the use of a novel kernel-based technique for combining modalities under fully general conditions, namely, the neutral-point method.This method presents a conservative kernel-based strategy for dealing with missing and disjoint training data in independent measurement modalities that can be theoretically shown to default to the Sum Rule classification scheme.Results of comparative experiments indicate that the neutral-point technique loses relatively little classification information with respect to coincident training data, and is in fact preferable for independent kernels produced by different physical modalities due to its better error-cancellation properties.

关键词： Kernel-based pattern recognition Combining modalities Kernel fusion Classifier fusion

作者: DAVID WINDRIDGE VADIM MOTTL ALEXANDER TATARCHLK ANDREY ELISEYEV

作者单位: Centre for Vision, Speech and Signal Processing, University of Surrey, The Stag Hill, Guildford, GU2 Computing Center of the Russian Academy of Sciences, Vavilov St.40, Moscow, 119991, Russia Moscow Institute of Physics and Technology, Institutsky Per.9, Dolgoprudny, Moscow Region, 141700, R

会议类型: 国际会议

会议名称: 2007 International Conference on Machine Learning and Cybernetics(IEEE第六届机器学习与控制论国际会议)

会议地点: 香港

会议语种:英文

页码: 3594-3600

在线出版日期: 2007-08-19（万方平台首次上网日期，不代表论文的发表时间）