Selection of significant input variables for time series forecasting |
| |
| Affiliation: | 1. School of Civil, Environmental and Chemical Engineering, RMIT University, GPO Box 2476, Melbourne, Victoria 3001, Australia;2. College of Engineering and Science, Victoria University, PO Box 14428, Melbourne, Victoria 8001, Australia;3. Institute for Sustainability and Innovation, Victoria University, PO Box 14428, Melbourne, Victoria 8001, Australia;1. College of Information Engineering, Zhejiang University of Technology, 288 Liuhe Road, Hangzhou, Zhejiang, China;2. Institute of Zhejiang ChinaOly Police Cloud Information, Hangzhou ChinaOly Tech. Co. Ltd., 9 Jiusheng Road, Hangzhou, Zhejiang, China;3. Graduate School of Information, Production and Systems, Waseda University, 2-7 Hibikino, Kitakyushu, Fukuoka, Japan;1. Bangor University, Dean Street, Bangor Gwynedd LL57 1UT, United Kingdom;2. University of Burgos, Escuela Politécnica Superior, Avda. de Cantabria s/n, 09006 Burgos, Spain;3. The University of Nottingham, NG8 1BB, United Kingdom;1. Melbourne School of Psychological Sciences, The University of Melbourne, Australia;2. Commonwealth Scientific & Industrial Research Organization, Australia;3. Faculty of Health, Arts and Design, Swinburne University of Technology, Australia |
| |
| Abstract: | Appropriate selection of inputs for time series forecasting models is important because it not only has the potential to improve performance of forecasting models, but also helps reducing cost in data collection. This paper presents an investigation of selection performance of three input selection techniques, which include two model-free techniques, partial linear correlation (PLC) and partial mutual information (PMI) and a model-based technique based on genetic programming (GP). Four hypothetical datasets and two real datasets were used to demonstrate the performance of the three techniques. The results suggested that the model-free PLC technique due to its computational simplicity and the model-based GP technique due to its ability to detect non-linear relationships (demonstrated by its relatively good performance on a hypothetical complex non-linear dataset) are recommended for the input selection task. Candidate inputs which are selected by both these recommended techniques should be considered as significant inputs. |
| |
| Keywords: | Time series forecasting Input variable selection Genetic programming Partial mutual information Correlation |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
