An empirical comparison of machine learning models for time series forecasting

• Published in: Econometric reviews Vol. 29; no. 5; pp. 594 - 621
• Main Authors: Ahmed, Nesreen K, Atiya, Amir F, Gayar, Neamat el, El-Shishiny, Hisham
• Format: Journal Article
• Language: English
• Published: New York Taylor & Francis Group 01-09-2010; Taylor and Francis Journals; Taylor & Francis Ltd
• Series: Econometric Reviews
• Subjects: Bayesian analysis; Comparative studies; Comparison study; Computational methods; Economic forecasts; Empirical analysis; Forecasting; Forecasting techniques; Gaussian; Gaussian process regression; Learning; Lernen; Machine learning; Machine learning models; Mathematical models; Multilayer perceptrons; Neural network forecasting; Neural networks; Neuronales Netz; Normal distribution; Preprocessing; Prognose; Regression; Regression analysis; Stochastic processes; Support vector regression; Time series; Verfahren; Zeitreihenanalyse
• ISSN: 0731-1761; 0747-4938; 1532-4168
• DOI: 10.1080/07474938.2010.481556

In this work we present a large scale comparison study for the major machine learning models for time series forecasting. Specifically, we apply the models on the monthly M3 time series competition data (around a thousand time series). There have been very few, if any, large scale comparison studies for machine learning models for the regression or the time series forecasting problems, so we hope this study would fill this gap. The models considered are multilayer perceptron, Bayesian neural networks, radial basis functions, generalized regression neural networks (also called kernel regression), K-nearest neighbor regression, CART regression trees, support vector regression, and Gaussian processes. The study reveals significant differences between the different methods. The best two methods turned out to be the multilayer perceptron and the Gaussian process regression. In addition to model comparisons, we have tested different preprocessing methods and have shown that they have different impacts on the performance.