一种提高文本分类效果的改进流形学习算法

流形学习算法在模式识别领域有着重要应用,针对文本分类数据的特点,提出一种基于邻域选取进行修正的局部线性嵌入算法,用带有权值的欧式距离来构造文本数据的局部邻域,提高文本分类的识别率;同时,利用文本数据的类别信息,运用半监督局部线性嵌入算法构造分类器,提高文本分类的效果。实验表明,本文基于文本分类改进的流形学习算法,能够有效地对文本进行分类。

An Improved Text Classification Results of the Improved Manifold Learning Algorithm

Manifold learning algorithms play an important part in the field of pattern recognition.In this paper,we propose a based on neighbors selecting modified Local Linear Embedding algorithm,which uses the weighted Euclidean distance to construct the local neighborhood of the text data to improve text classification recognition rate.At the same time,we use the Semi-Supervised Locally Linear Embedding algorithm to improve the text classification.Experiments show that the text-based manifold learning algorithms th...