ِِِThe Comparison of Adaptive Filtering and Back Propagation Neural Networks in Time Series

. Abstract An increasing interest in studying the time series has been recently regarded. One of the potential methods in analyzing the time series and forecasting the future values is the adaptive filtering. A modern method of analyzing time series is the artificial networks in forecasting. The purpose of current paper is to compare the adaptive filtering ARMA(p,q) model with Back propagation neural networks. The mean square error, mean absolute error and mean absolute percentage has been adopted as a statistical criteria mean to find the best one and get a good result. Application was done on alive births of males' data of Mosul city using ANNs, and AF techniques.