Application of Rank Likelihood Ratio Scanning Method in Multiple Mean Changes in Long Memory Time Series with Heavy Tail

Abstract

In this paper, a rank-likelihood ratio scanning method for long memory time series with heavy tail is proposed to solve the problem that when likelihood ratio scanning method is used to estimate the mean change points in the long memory time series with heavy tail, the estimation accuracy decreases rapidly with the decrease of the heavy tail index. Numerical simulation and analysis of real data demonstrate the effectiveness and practicability of the rank-likelihood ratio scanning method.