TIME SERIES FORECASTING: COMPARISON OF BOX-JENKINS, HYBRID AND MACHINE LEARNING METHODS

Details

Supervisors

von Sachs, Rainer

Faculty

Faculté des sciences

Degree label

Master [120] en science des données, orientation statistique, à finalité spécialisée

Abstract

This thesis mainly discusses forecasting competitions and how they contribute to the field of forecasting. It also discusses the winner of the M4 competition, a hybrid model that blends statistical and deep learning concepts. The model, called Exponential Smoothing Recurrent Neural Network (ES-RNN), was proposed by Slawek Smyl, a data scientist at Uber. This hybrid model outperformed all other models in the M4 competition, pointing the way for further development and research into more complex methods.