搜索资源列表
CLIQUE
- CLIQUE(Clustering In QUEst)是一种简单的基于网格的聚类方法,用于发现子空间中基于密度的簇。CLIQUE把每个维划分成不重叠的区间,从而把数据对象的整个嵌入空间划分成单元。它使用一个密度阈值识别稠密单元和稀疏单元。一个单元是稠密的,如果映射到它的对象数超过该密度阈值。(CLIQUE (Clustering In QUEst) is a simple grid based clustering method for the discovery of clusters bas
Kmeans
- 算法思想:提取文档的TF/IDF权重,然后用余弦定理计算两个多维向量的距离来计算两篇文档的相似度,用标准的k-means算法就可以实现文本聚类。源码为java实现(Algorithm idea: extract the TF/IDF weight of the document, then calculate the distance between two multidimensional vectors by cosine theorem, calculate the similarity
clique
- 使用java为开发语言,基于高维空间网格聚类算法CLIQUE(Use java as the development language, based on high dimensional space grid clustering algorithm CLIQUE)
Word2VEC_java-master
- word2vec包含训练模型,聚类,可以结合ansj分词使用(Word2vec contains training models, and clustering can be combined with ansj participle.)