Skeleton Decomposition Analysis for Subspace Clustering

2016-12-08
Sekmen, Ali
Aldroubi, Akram
Koku, Ahmet Buğra
This paper provides a comprehensive analysis of skeleton decomposition used for segmentation of data W = [w(1) center dot center dot center dot w(N)] subset of R-D drawn from a union u = U-i=1(M) S-i of linearly independent subspaces {Si}(M)(i=1) of dimensionsof {di}(M)(i=1). Our previous work developed a generalized theoretical framework for computing similarity matrices by matrix factorization. Skeleton decomposition is a special case of this general theory. First, a square sub-matrix A is an element of R-rxr of W with the same rank r as W is found. Then, the corresponding row restriction R of W is constructed. This leads to P= A(-1)R and corresponding similarity matrix SW = ((PP)-P-T)(dmax), where d(max) is the maximum subspace dimension. Since most of the data matrices are low-rank in many subspace segmentation problems, this is computationally efficient compared to the other constructions of similarity matrices. It is also shown (with some limitations) that center-of-mass based sorting of data columns in SW can be used to quickly assess clustering performance while algorithm development in both noisy or noise-free cases.

Suggestions

Similarity matrix framework for data from union of subspaces
Aldroubi, Akram; Sekmen, Ali; Koku, Ahmet Buğra; Cakmak, Ahmet Faruk (2018-09-01)
This paper presents a framework for finding similarity matrices for the segmentation of data W = [w(1)...w(N)] subset of R-D drawn from a union U = boolean OR(M)(i=1) S-i, of independent subspaces {S-i}(i=1)(M), of dimensions {d(i)}(i=1)(M). It is shown that any factorization of W = BP, where columns of B form a basis for data W and they also come from U, can be used to produce a similarity matrix Xi w. In other words, Xi w(i, j) not equal 0, when the columns w(i) and w(j) of W come from the same subspace, ...
Matrix resconstruction: Skeleton decomposition versus singular value decomposition
SEKMEN, ali; ALDROUBİ, Akram; Koku, Ahmet Buğra; HAMM, Keaton (2017-07-12)
In this work, Skeleton Decomposition (SD) and Singular Value Decomposition (SVD) are compared and evaluated for reconstruction of data matrices whose columns come from a union of subspaces. Specifically, an original data matrix is reconstructed from noise-contaminated version of it. First, matrix reconstruction using SD iteratively is introduced and alternative methods for forming SD-based reconstruction are discussed. Then, through exhaustive simulations, effects of process parameters such as noise level, ...
Principal Coordinate Clustering
SEKMEN, ali; ALDROUBİ, Akram; HAMM, Keaton; Koku, Ahmet Buğra (2017-12-14)
This paper introduces a clustering algorithm, called principal coordinate clustering. It takes in a similarity matrix SW of a data matrix W and computes the singular value decomposition of SW to determine the principal coordinates to convert the clustering problem to a simpler domain. It is a relative of spectral clustering, however, principal coordinate clustering is easier to interpret, and gives a clear understanding of why it performs well. In a fashion, this gives intuition behind why spectral clusteri...
Physical subspace identification for helicopters
Avcıoğlu, Sevil; Kutay, Ali Türker; Department of Aerospace Engineering (2019)
Subspace identification is a powerful tool due to its well-understood techniques based on linear algebra (orthogonal projections and intersections of subspaces) and numerical methods like QR and singular value decomposition. However, the state space model matrices which are obtained from conventional subspace identification algorithms are not necessarily associated with the physical states. This can be an important deficiency when physical parameter estimation is essential. This holds for the area of helico...
Three dimensional hyperbolic grid generation
Dinçgez, Umut Can; Aksel, Mehmet Haluk; Department of Mechanical Engineering (2006)
This thesis analyzes procedure of generation of hyperbolic grids formulated by two constraints, which specify grid orthogonality and cell volume. The procedure was applied on a wide range of geometries and high quality two and three dimensional hyperbolic grids were generated by using grid control and smoothing procedures, which supply grid clustering in all directions and prevent grid deformation (grid shock), respectively.
Citation Formats
A. Sekmen, A. Aldroubi, and A. B. Koku, “Skeleton Decomposition Analysis for Subspace Clustering,” 2016, Accessed: 00, 2020. [Online]. Available: https://hdl.handle.net/11511/41679.