Time series prediction using Artificial Neural Network

Abstract

This study investigated the use of Multilayer Perceptron (MLP) artificial neural network and Autoregressive Integrated Moving Average (ARIMA) models for time series prediction. The models are evaluated using two statistical performance evaluation measures, Root Mean Squared Error (RMSE) and coefficient of determination (R²). Four different multilayer perceptron were developed and compared. The best MLP was then compared with ARIMA model. The experimental result shows that a 3-layer MLP architecture using tanh activation function in each of the hidden layer and linear function in the output layer has the lowest prediction error and the highest coefficient of determination among the configured multilayer perceptron neural network. Comparative analysis of performance result reveals that multilayer perceptron neural network MLP has a lower prediction error than ARIMA model.

Keywords: Artificial Neural Network, ARIMA, Multilayer Perceptron, Time Series, Data Preprocessing