Adaptive Filtering: An Integrated Autoregressive/Moving Average Filter for Time Series Forecasting

This paper extends the applicability of a heuristic filtering technique, adaptive filtering, by dealing with a number of practical considerations in time series forecasting. These are problems that have been raised by other researchers examining this technique and by practitioners using it for time series analysis. These modifications make adaptive filtering much more comparable to the Box-Jenkins methodology for autoregressive/moving average processes. A specific application of adaptive filtering is provided.