103High-Performance Tool for the Test of Long-Memory and Self-Similarity
4.4.1 selfis
Sels (Self-similarity analysis) is a java-based self-similarity analysis software tool developed by
Thomas Karagiannis at University of California at Riverside. Current algorithms implemented in
Sels include R/S statistic, variance type, absolute moment, variance of residuals, local Whittle,
periodogram, and Abry–Veitch method. Additionally, Sels performs basic statistics computations,
ACF, and power spectral density estimation, and some cleansing algorithms such as bucket shuf-
ing. Sels functionality is the same as in SelQoS. First, for Hurst-index estimation, the opening
of a one-column le is selected; after this, the time series is plotted and sho ...