How Much Information Does Dependence Between Wavelet Coefficients Contain

摘要: ABSTRACTThis article is motivated by several articles that propose statistical inference where the independence of wavelet coefficients for both short- as well long-range dependent time series assumed. We focus on sample variance and investigate influence dependence between this statistic. To end, we derive asymptotic distributional properties a synthesized, ignoring some or all coefficients. show second-order differ from those true whose have same marginal distribution except in independent Gaussian case. This holds even if dependency correct within each level only levels ignored. In case autocovariances autocorrelations at lag one, indicate first-order are erroneous. second step, nonparametric b...