Outlier detection and localisation with wavelet based multifractal formalism

摘要: We present a method of detecting and localising outliers in stochastic processes. The checks the internal consistency scaling behaviour process within paradigm multifractal spectrum. Deviation from expected spectrum is interpreted as potential presence outliers. detection part then supplemented by localisation analysis part, using local properties time series. Localised can be removed one one, with possibility dynamic verification spectral properties. Both formalism series are implemented on wavelet transform modulus maxima tree.