基于矩阵化的多视角分类学习算法
首发时间:2010-05-14
摘要:本文从原始单视角的向量模式衍生出多种多视角的矩阵表示,从而基于新产生的多种矩阵表示设计出一种新的多视角分类算法。具体过程是,首先将原始单视角模式的向量表示重组成多种矩阵表示形式,接着再将一种给定的分类器重构成能处理多种矩阵模式表示的多个分类器,将每个新生成的分类器视为对原单视角模式一个新视角下的学习,从而形成了一组对原单视角模式有新描述的多视角分类器算法,最后对这些生成的多个视角下的分类器采用一联合而非分离的学习过程。在实践中,本文采用向量型正则化Ho-Kashyap分类器模型(Modification of Ho-Kashyap algorithm with Squared approximation of the misclassification errors, MHKS)作为基分类器,根据不同的矩阵表示形成多个矩阵型的正则化Ho-Kashyap分类算法(Matrix-pattern-oriented MHKS, MatMHKS),最后对多个MatMHKS算法联合学习从而形成了一种新的多视角分类算法(MultiV-MHKS)。实验验证了所提算法的可行性与有效性。
关键词: 模式表示 矩阵模式 向量模式 分类器设计 多视角学习
For information in English, please click here
A Novel Multi-view Classifier Based on Matrixization
Abstract:This paper aims at developing a new MVL technique for single source patterns. First, we reshape the original vector representation of single source patterns into multiple matrix representations. Doing so can change the original architecture of a given base classifier into different sub-ones. Each newly-generated sub-classifier can classify the patterns represented with the matrix, and is taken as one view of the original base classifier. As a result, a set of sub-classifiers with different views are come into being. Then, a joint rather than separated learning process for the multi-view sub-classifiers is developed. In practice, the original base classifier employs the vector-pattern-oriented Ho-Kashyap classifier with regularization learning (called MHKS) as a paradigm that is not limited to MHKS. Finally, the feasibility and effectiveness of the proposed multi-view-combined classifier, named MultiV-MHKS, is demonstrated by the experimental results on benchmark data sets.
Keywords: Pattern Representation Matrix Pattern Vector Pattern Classifier design Multi-view learning
基金:
论文图表:
引用
No.3585545305071273****
同行评议
勘误表
基于矩阵化的多视角分类学习算法
评论
全部评论0/1000