A multi-method forecasting algorithm: Linear unbiased estimation of combine forecast

• Published in: Knowledge-based systems Vol. 239; p. 107990
• Main Authors: Bekiroglu, Korkut, Gulay, Emrah, Duru, Okan
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 05.03.2022; Elsevier Science Ltd
• Subjects: Algorithms; Combine forecasting; Crude oil prices; Forecasting; Methodology; Model learning; Performance prediction; Sparse model combination; Sparse model combination; Combine forecasting; Model learning

This paper proposes a dynamic ensemble algorithm to combine forecasting results from multiple methodologies subject to their local (recent) predictive performance. In contrast to conventional combination forecasts, the proposed algorithm runs a sparsification process to merge a subset of methodology...