Inscriere cercetatori

Daca aveti cont Ad Astra si de Facebook, intrati pe pagina de profil pentru a da dreptul sa va logati pe site doar cu acest buton.

Site nou !

Daca nu va puteti recupera parola (sau aveti alte probleme), scrieti-ne la pagina de contact. Situl vechi se gaseste la adresa


Properties of multivariate data investigated by fractal dimensionality

Domenii publicaţii > Biologie + Tipuri publicaţii > Articol în revistã ştiinţificã

Autori: Nikolic D, Moca VV, Singer W and Muresan RC

Editorial: Journal of Neuroscience Methods, 172(1), p.27-33, 2008.


Elaborated data-mining techniques are widely available today. Nevertheless, many non-linear relations among variables remain undiscovered in multi-dimensional datasets. To address this issue we propose a method based on the concept of fractal dimension that explores the structure of multivariate data and apply the method to simulated data, as well as to local field potentials recorded from cat visual cortex. We find that with changes in the analysis scale, the dimensionality of the data often changes, indicating first that the data are not simple fractals with one unique dimension and second, that, at a certain scale, important changes in the geometric structure of the data may occur. The method can be used as a data-mining tool but also as a method for testing a model’s fit to the data. We achieve the latter by comparing the dimensionality of the original data to the dimensionality of the data reconstructed from a model’s description of the data (here using the general linear model). The method provides indispensable help in estimating the complexity of non-linear relationships within multivariate datasets.

Cuvinte cheie: Fractal; Data model; Multivariate analysis; Dimensionality