OPTICS
Sign in to savealgorithm for finding density based clusters in spatial data
Article · 中文
OPTICS(英語:Ordering points to identify the clustering structure)是由Mihael Ankerst,Markus M. Breunig,Hans-Peter Kriegel和Jörg Sander提出的基于密度的聚类分析算法。OPTICS并不依赖全局变量来确定聚类,而是将空间上最接近的点相邻排列,以得到数据集合中的对象的线性排序。排序后生成的序列存储了与相邻点之间的距离,并最终生成了一个 dendrogram 。OPTICS算法的思路与DBSCAN类似,但是解决了DBSCAN的一个主要弱点,即如何在密度变化的数据中取得有效的聚类。同时 OPTICS也避免了多数聚类算法中对输入参数敏感的问题。
Abstract from DBpedia / Wikipedia · CC BY-SA
Connections
artificial neural network
Entity
online machine learning
Entity
International Standard Book Number
Entity
digital object identifier
Entity
machine learning
Entity
R
Entity
deep learning
Entity
data mining
Entity
regression analysis
Entity
crowdsourcing
Entity
supervised learning
Entity
cluster analysis
Entity
reinforcement learning
Entity
linear regression
Entity
heap
Entity
factor analysis
Entity
Q22908627
Entity
principal component analysis
Entity
transformer
Entity
unsupervised learning
Entity