The ARMA method of approximating probability density functions

From MaRDI portal





A new method of approximating probability density functions is proposed. The method uses rational representing functions of the spectrum of an autoregressive moving average (ARMA) time series. Conditions are given under which certain ARMA approximators have essentially smaller integrated squared errors than truncated Fourier series approximators. Several examples illustrating the improved performance of ARMA, over Fourier series, approximators are considered.











This page was built for publication: The ARMA method of approximating probability density functions

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1068482)