Green Clustered mean Forecasting (ECMF) for Time series is a novel forecasting approach that is capable of efficaciously aggregating the underlying additives of a time collection through combining the top-primarily based Aggregation (HBA) technique alongside the mean Forecasting (MF) method. ECMF may be utilized to detect and become aware of companies of time collection intently associated with every other, therefore taking into account extra correct and reliable forecasts. By way of counting on HBA to discover any patterns in the information and then imposing MF to generate forecasts from the ones styles, ECMF can offer more accurate and reliable forecasts than the prevailing strategies. Moreover, this technique is computationally green, due to the hierarchical shape of the data, which reduces the quantity of forecasting experiments wanted. The ensuing forecasts are also easier to examine, as they may be generated from a standard head-based totally version.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Efficient Clustered Mean Forecasting for Time Series with Head-Based Aggregation

  • Piyush Anand,
  • T. Ramesh,
  • Trapty Agarwal,
  • D. Preethi

摘要

Green Clustered mean Forecasting (ECMF) for Time series is a novel forecasting approach that is capable of efficaciously aggregating the underlying additives of a time collection through combining the top-primarily based Aggregation (HBA) technique alongside the mean Forecasting (MF) method. ECMF may be utilized to detect and become aware of companies of time collection intently associated with every other, therefore taking into account extra correct and reliable forecasts. By way of counting on HBA to discover any patterns in the information and then imposing MF to generate forecasts from the ones styles, ECMF can offer more accurate and reliable forecasts than the prevailing strategies. Moreover, this technique is computationally green, due to the hierarchical shape of the data, which reduces the quantity of forecasting experiments wanted. The ensuing forecasts are also easier to examine, as they may be generated from a standard head-based totally version.