On the reduction of Gaussian inverse Wishart mixtures

Date

2012-09-12

Author

Granström, Karl
Orguner, Umut

Metadata

Show full item record

Item Usage Stats

249
views

0
downloads

This paper presents an algorithm for reduction of Gaussian inverse Wishart mixtures. Sums of an arbitrary number of mixture components are approximated with single components by analytically minimizing the Kullback-Leibler divergence. The Kullback-Leibler difference is used as a criterion for deciding whether or not two components should be merged, and a simple reduction algorithm is given. The reduction algorithm is tested in simulation examples in both one and two dimensions. The results presented in the paper are useful in extended target tracking using the random matrix framework.

Subject Keywords

Merging, Target tracking, Approximation methods, Approximation algorithms, Signal processing algorithms, Covariance matrix, Symmetric matrices

URI

https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84867656716&origin=inward
https://hdl.handle.net/11511/73950
https://ieeexplore.ieee.org/document/6290566

Conference Name

15th International Conference on Information Fusion - FUSION 2012 (2012)

Collections

Department of Electrical and Electronics Engineering, Conference / Seminar

Suggestions

OpenMETU
Core

On the Poisson sum formula for the analysis of wave radiation and scattering from large finite arrays Aydın Çivi, Hatice Özlem; Chou, HT (1999-05-01) Poisson sum formulas have been previously presented and utilized in the literature [1]-[8] for converting a finite element-by-element array field summation into an alternative representation that exhibits improved convergence properties with a view toward more efficiently analyzing wave radiation/scattering from electrically large finite periodic arrays. However, different authors [1]-[6] appear to use two different versions of the Poisson sum formula; one of these explicitly shows the end-point discontinui...
Improving the big bang-big crunch algorithm for optimum design of steel frames Hasançebi, Oğuzhan (null; 2012-01-01) This paper presents an improved version of the big bang-big crunch (BB-BC) algorithm namely exponential BB-BC algorithm (EBB-BC) for optimum design of steel frames according to ASD-AISC provisions. It is shown that the standard version of the algorithm sometimes is unable to provide reasonable solutions for problems from discrete design optimization of steel frames. Therefore, by investigating the shortcomings of the BB-BC algorithm, it is aimed to enhance the algorithm for solving complicated steel frame o...
On the arithmetic operations over finite fields of characteristic three with low complexity AKLEYLEK, SEDAT; Özbudak, Ferruh; Özel, Claire Susanna (2014-03-15) In this paper, the Hermite polynomial representation is adapted as a new way to represent certain finite fields of characteristic three. We give the multiplication method to multiply two elements of F-3n in the Hermite polynomial representation with subquadratic computational complexity by using a divide-and-conquer idea. We show that in some cases there is a set of irreducible binomials in the Hermite polynomial representation to obtain modular reduction with a lower addition complexity than the standard p...
On the generalisation of special moduli for faster interleaved montgomery modular multiplication AKLEYLEK, SEDAT; Cenk, Murat; Özbudak, Ferruh (2013-09-01) In this study, the authors give a generalisation of special moduli for faster interleaved Montgomery modular multiplication algorithm with simplified pre-computational phase for GF(p(n)), where p 2 is a prime number and n is a positive integer. The authors propose different sets of moduli that can be used in elliptic curve crytographic applications and pairing-based cryptography. Moreover, this method also leads to efficient implementations for the elliptic curve parameters given in standards. It is shown t...
A study on transmit beamforming and noisy autoregressive modeling problems Çayır, Ömer; Candan, Çağatay; Department of Electrical and Electronics Engineering (2023-1) This thesis presents algorithms for two problems in statistical signal processing. The first problem is the transmit beamformer design under the peak-to-average power ratio (PAPR) constraint. With the aim of establishing a trade-off between the power efficiency (maximizing the average transmitted power in the main lobe) and other metrics, such as the fluctuation of power in the main lobe and the peak-sidelobe level, how the PAPR constraint affects the design problem is examined. In general, unimodular weigh...

Citation Formats

K. Granström and U. Orguner, “On the reduction of Gaussian inverse Wishart mixtures,” Singapur, 2012, p. 2162, Accessed: 00, 2021. [Online]. Available: https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84867656716&origin=inward.