rian.ie - Browse

Author(s)
Institution
Publication types
Funder
Year

Limited By:	Subject = Transforms;

3 items found

Sort by

Order

per page

Displaying Results 1 - 3 of 3 on page 1 of 1

Marked Mark

Convolutive non-negative matrix factorisation with a sparseness constraint (2006)

Pearlmutter, Barak A.; O'Grady, Paul D.

Convolutive non-negative matrix factorisation with a sparseness constraint (2006)

Pearlmutter, Barak A.; O'Grady, Paul D.

Abstract:

Discovering a representation which allows auditory data to be parsimoniously represented is useful for many machine learning and signal processing tasks. Such a representation can be constructed by non-negative matrix factorisation (NMF), a method for finding parts-based representations of non-negative data. We present an extension to NMF that is convolutive and includes a sparseness constraint. In combination with a spectral magnitude transform, this method discovers auditory objects and their associated sparse activation patterns.

http://mural.maynoothuniversity.ie/1375/

Marked Mark

Fukunaga-Koontz transform for small sample size problems (2005)

Miranda, Abhilash A.; Whelan, Paul F.

Fukunaga-Koontz transform for small sample size problems (2005)

Miranda, Abhilash A.; Whelan, Paul F.

Abstract:

In this paper, we propose the Fukunaga-Koontz Transform (FKT) as applied to small-sample size (SSS) problems and formulate a feature scatter matrix based equivalent of the FKT. We establish the classical linear discriminant analysis (LDA) analogy of the FKT and apply it to a SSS situation. We demonstrate the significant computational savings and robustness associated with our approach using a multi-class face detection problem

http://doras.dcu.ie/4663/

Marked Mark

Preservation of Common Quadratic Lyapunov Functions and Padé Approximations (2010)

Sajja, Surya; Solmaz, Selim; Shorten, Robert N.; Corless, Martin

Preservation of Common Quadratic Lyapunov Functions and Padé Approximations (2010)

Sajja, Surya; Solmaz, Selim; Shorten, Robert N.; Corless, Martin

Abstract:

It is well known that the bilinear transform, or first order diagonal Padé approximation to the matrix exponential, preserves quadratic Lyapunov functions between continuous-time and corresponding discrete-time linear time invariant (LTI) systems, regardless of the sampling time. It is also well known that this mapping preserves common quadratic Lyapunov functions between continuous-time and discrete-time switched systems. In this note we show that while diagonal Padé approximations do not in general preserve other types of Lyapunov functions (or even stability), it is true that diagonal Padé approximations of the matrix exponential, of any order and sampling time, preserve quadratic stability. A consequence of this result is that the quadratic stability of switched systems is robust with respect to certain discretization methods.

http://mural.maynoothuniversity.ie/3826/

Displaying Results 1 - 3 of 3 on page 1 of 1