Automatic forecasting for univariate time series with long memory property

In applications of time series forecasting, it is necessary to model the data automatically and obtain the corresponding forecast values, possibly in real-time, using various methods. A number of automatic algorithms for modeling univariate time series data are available in the literature. This paper is restricted to a study on automatic forecasting algorithms for time series data which contain long memory property using automatic Fractional Autoregressive Integrated Moving Average (FARIMA/ARFIMA) model. The algorithms are implemented through open-source software R. We provide empirical application to illustrate and show applicability of the proposed methods and tools using real data