Time Series Forecasting Using Hybrid ARIMA and ANN Models Based on DWT Decomposition

• DOI: 10.1016/j.procs.2015.04.167
• Publication Date: 2015-05-22T06:04:21Z
• Authors: Ina Khandelwal; Ratnadip Adhikari; Ghanshyam Verma
• Keywords: Moving average

Recently Discrete Wavelet Transform (DWT) has led to a tremendous surge in many domains of science and engineering. In this study, we present the advantage DWT improve time series forecasting precision. This article suggests novel technique by segregating dataset into linear nonlinear components through DWT. At first, is used decompose in-sample training (detailed) non-linear (approximate) parts. Then, Autoregressive Integrated Moving Average (ARIMA) Artificial Neural Network (ANN) models are separately recognize predict reconstructed detailed approximate components, respectively. manner, proposed approach tactically utilizes unique strengths DWT, ARIMA, ANN accuracy. Our hybrid method tested on four real-world its results compared with those ANN, Zhang's models. Results clearly show that achieves best accuracies for each series.