Exécution de notre outil de synthèse sur un nouvel article

Résumé : This paper introduces a new graphical method of displaying the evolutionary correlation properties of nonstationary time series. The method, called time‐correlation analysis (TCA), summarizes the local correlation properties with a smooth and monotone characterization function interpretable as the first‐order autocorrelation coefficient. The TCA plot is constructed by plotting the discretized characterization function as time series. Issues concerning the range of the TCA plot are investigated in detail. Both simulated and real‐data examples are given to demonstrate the application and interpretation of the method.