Abstract
Forecasting factory productivity is a critical task. However, it is not easy owing to the uncertainty of productivity. Existing methods often forecast productivity using a fuzzy number. However, the range of a fuzzy productivity forecast is wide owing to the consideration of extreme cases. In this study, a fuzzy collaborative forecasting approach is proposed to forecast factory productivity using a type-II fuzzy number and by narrowing the forecast's range. The outer section of the type-II fuzzy number determines the range of productivity, while the inner section is defuzzified to derive the most likely value. Based on the experimental results, the proposed methodology surpassed existing methods in improving forecasting precision and accuracy, with a reduction in the mean absolute percentage error (MAPE) of up to 74%.
Original language | English |
---|---|
Number of pages | 13 |
Journal | Journal of Ambient Intelligence and Humanized Computing |
DOIs | |
State | E-pub ahead of print - 5 Aug 2020 |
Keywords
- Fuzzy collaborative forecasting
- Type-II fuzzy number
- Mixed binary nonlinear programming
- Productivity
- INTELLIGENCE APPROACH
- LINEAR-REGRESSION