Multiple change point detection by sparse parameter estimation

Warning

This publication doesn't include Institute of Computer Science. It includes Faculty of Economics and Administration. Official publication website can be found on muni.cz.

Authors	NEUBAUER Jiří VESELÝ Vítězslav
Year of publication	2010
Type	Article in Proceedings
Conference	Proceedings of COMPSTAT'2010, 19th International Conference on Computational Statistics
MU Faculty or unit	Faculty of Economics and Administration
Citation
Web	http://www.econ.muni.cz/~vesely/papers/Compstat10.pdf
Field	Applied statistics, operation research
Keywords	multiple change point detection; overparametrized model; sparse parameter estimation; basis pursuit algorithm
Description	The contribution is focused on multiple change point detection in a onedimensional stochastic process by sparse parameter estimation from an overparametrized model. Stochastic process with changes in the mean is estimated using dictionary consisting of Heaviside functions. The basis pursuit algorithm is used to get sparse parameter estimates. Some properties of mentioned method are studied by simulations.
Related projects:	Dynamic Geovisualisation in Crises Management