Ensemble methods for advanced skier days prediction
作者：
Highlights：
• We propose the use of ensemble learning techniques to improve skier days forecasts.
• We utilize actual data containing skier days across six ski seasons, from a regional ski resort.
• Three base prediction models as inputs to four ensemble modeling techniques are analyzed.
• Stacking shows the highest increase in accuracy across five independent variable versions.
• An operational prediction model is offered.
摘要
Highlights•We propose the use of ensemble learning techniques to improve skier days forecasts.•We utilize actual data containing skier days across six ski seasons, from a regional ski resort.•Three base prediction models as inputs to four ensemble modeling techniques are analyzed.•Stacking shows the highest increase in accuracy across five independent variable versions.•An operational prediction model is offered.
论文关键词：Ensemble learning,Data mining,Forecasting,Skier days
论文评审过程：Available online 14 August 2013.
论文官网地址：https://doi.org/10.1016/j.eswa.2013.08.002
