Lecture 15 : Spectral clustering , projective clustering
نویسندگان
چکیده
Figure 15.1: k=3 clusters with red points chosen as facilities. Consider a situation where we have n point locations and we wish to place k facilities among these points to provide some service. It is desirable to have these facilities close to the points they are serving, but the notion of “close” can have different interpretations. The k-means problem seeks to place k facilities so as to minimize the variance between points and the facilities they are assigned to. The k-median problem is similar, but seeks to minimize the average distance of facility locations to points.
منابع مشابه
Projective Low-rank Subspace Clustering via Learning Deep Encoder
Low-rank subspace clustering (LRSC) has been considered as the state-of-the-art method on small datasets. LRSC constructs a desired similarity graph by low-rank representation (LRR), and employs a spectral clustering to segment the data samples. However, effectively applying LRSC into clustering big data becomes a challenge because both LRR and spectral clustering suffer from high computational...
متن کاملProjective Clustering Method for the Detection of Outliers in Non-Axis Aligned Subspaces
Clustering the case of non-axis-aligned subspaces and detection of outliers is a major challenge due to the curse of dimensionality. The normal clustering was efficient in axis-aligned subspaces only. To solve this problem, projective clustering has been defined as an extension to traditional clustering that attempts to find projected clusters in subsets of the dimensions of a data space. A pro...
متن کاملA Review of Nonnegative Matrix Factorization Methods for Clustering
Nonnegative Matrix Factorization (NMF) was first introduced as a low-rank matrix approximation technique, and has enjoyed a wide area of applications. Although NMF does not seem related to the clustering problem at first, it was shown that they are closely linked. In this report, we provide a gentle introduction to clustering and NMF before reviewing the theoretical relationship between them. W...
متن کاملLecture 7 — April 21 7.1 Spectral Clustering
is the degree matrix and elucidate some of its key properties that make it suitable for clustering. Suppose we form a weighted graph G with our data-points being the vertices and the edge weights being specified by the weight matrix W . Let us first consider the special case when the graph G has exactly k components with vertex sets A1, · · · , Ak. Then the Laplacian Lrw has the following prope...
متن کاملClustering cancer gene expression data by projective clustering ensemble
Gene expression data analysis has paramount implications for gene treatments, cancer diagnosis and other domains. Clustering is an important and promising tool to analyze gene expression data. Gene expression data is often characterized by a large amount of genes but with limited samples, thus various projective clustering techniques and ensemble techniques have been suggested to combat with th...
متن کامل