摘要
SVM算法复杂度与样本维数无关,具有的泛化能力强、分类精度高的特点,而LLE是有效的非线性降维方法,本文利用支持向量机(SVM)算法对局域线性嵌入(LLE)算法进行改进,有效地解决了基于内容的图像检索中的高维特征向量的降维问题,实验表明具有较高的查全率和查准率。
SVM has the generalization ability and high precision classification, and the algorithm complexity has nothing to do with the dimension of samples, LLE is an effective non-linear dimension reduction methods. This paper improved Locally Linear Embedding （LLE） algorithm used with Support Vector Machine （SVM） algorithm, and solved the dimension reduction problems of high-dimensional characteristics vector in the content-based image retrieval. The experiments showed that has high accurate rate and complete rate.
出处
《齐齐哈尔大学学报（自然科学版）》
2009年第4期14-17,共4页
Journal of Qiqihar University(Natural Science Edition)
基金
黑龙江省教育厅项目(项目编号:11531419)
关键词
支持向量机
局域线性嵌入
图像检索
support vector machine
locally linear embedding
image retrieval
