首页 | 官方网站   微博 | 高级检索  
     

稀疏低秩双线性判别模型及其应用
引用本文:蒋琳,谭晓阳,刘俊.稀疏低秩双线性判别模型及其应用[J].计算机科学与探索,2010,4(7):654-661.
作者姓名:蒋琳  谭晓阳  刘俊
作者单位:南京航空航天大学,信息科学与技术学院,南京,210016
基金项目:国家自然科学基金,江苏省自然科学基金
摘    要:"高维度小样本"问题是模式识别应用中的主要障碍之一。跨越这一障碍的有效方法之一是采用参数矩阵的低秩逼近,目的是控制模型复杂度。常用的低秩逼近方法需要预先指定目标矩阵秩的大小(如主成分分析)。提出了一种新的基于稀疏约束的低秩判别模型,此模型通过对目标参数进行矩阵分解,然后分别对子成分施加低秩(稀疏)约束,从而达到低秩逼近的目的。进一步将这一思想嵌入一个双边判别模型,并用坐标下降法对目标函数进行优化,使得算法在低秩逼近的同时还有效利用了输入数据的空间特性,从而得到更好的推广性能。其有效性在一个安全生物识别应用上得到了验证。

关 键 词:稀疏低秩逼近  双边判别框架  主成分分析
修稿时间: 

Sparse Low Rank Bilinear Discriminative Model and its Application
JIANG Lin,TAN Xiaoyang,LIU Jun.Sparse Low Rank Bilinear Discriminative Model and its Application[J].Journal of Frontier of Computer Science and Technology,2010,4(7):654-661.
Authors:JIANG Lin  TAN Xiaoyang  LIU Jun
Affiliation:College of Computer Science & Technology, Nanjing University of Aeronautics & Astronautics, Nanjing 210016, China
Abstract:“High dimensionality and small size samples” is widely encountered in many real world machine learning applications. Low rank approximation to parametric matrix has recently been proven to be an effective method to control the complexity of models. Most of the previous low rank methods are required to specify the target rank by hand beforehand (e.g., principal component analysis), however, imposing the sparsity constraints on the parametric matrix can avoid this. In particular, under a bilinear discriminative framework, decomposing the parametric matrix and simultaneously constraining their ranks with the sparsity-inducing regularization will perform well. The result-ing problem can be efficiently solved with coordinate descent. This method is able to take the spatial information of structured data into account, leading to improved generalization capability. The feasibility and effectiveness of the proposed method is demonstrated on a security biometric application.
Keywords:sparse low rank approximation  bilinear discriminative framework  principal component analysis
本文献已被 维普 万方数据 等数据库收录!
点击此处可从《计算机科学与探索》浏览原始摘要信息
点击此处可从《计算机科学与探索》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号