RT book part
T1 Robust wavelet-based assessment of scaling with applications
A1 Hamilton, Erin
AD  K.
A1 Jeon, Seonghye
A1 Ramírez Cobo, Josefa
A1 Sky Lee, Kichun
A1 Hinton, Raymond J.
A1 Vidakovic, Brani
A2 Estadística e Investigación Operativa
K1 self-similarity, Hurst-exponent, Fractional Brownian Motion, 2D discrete wavelet transform, Robust Statistics
AB A number of approaches have dealt with statistical assessment of selfsimilarity, and many of those are based on multiscale concepts. Most rely on certain distributional assumptions which are usually violated by real data traces, often characterized by large temporal or spatial mean level shifts, missing values or extreme observations. A novel, robust approach based on Theil-type weighted regression is proposed for estimating self-similarity in two-dimensional data (images).The method is compared to two traditional estimation techniques that use waveletdecompositions; ordinary least squares (OLS) and Abry-Veitch bias correcting estimator (AV). As an application, the suitability of the self-similarity estimate resultingfrom the the robust approach is illustrated as a predictive feature in the classificationof digitized mammogram images as cancerous or non-cancerous. The diagnosticemployed here is based on the properties of image backgrounds, which is typicallyan unused modality in breast cancer screening. Classification results show nearly68% accuracy, varying slightly with the choice of wavelet basis, and the range ofmultiresolution levels used.
YR 2024
FD 2024
LK http://hdl.handle.net/10498/33504
UL http://hdl.handle.net/10498/33504
LA eng
DS Repositorio Institucional de la Universidad de Cádiz
RD 08-nov-2026