多方法联合预测"四川科技兴村"平台有效数据量
Forecast valid data of SiChuan-TechRural-Revitalization-Plat based on multi-method integration
廖建雄 1刘黎立 1程小帆2
作者信息
- 1. 四川博大纳川科技有限责任公司,四川 成都 610041
- 2. 四川省计算机研究院,四川 成都 610041
- 折叠
摘要
"四川科技兴村"平台有效数据量预测依赖于现有的移动平均法,而移动平均法在实际应用中无法准确捕捉数据的实时动态变化,导致预测结果未能达到期望的精度和效果.为了更准确地预测"四川科技兴村"平台有效数据量,提出了环同期预测法、次级平台加权平均法和数据留存评估法联合预测的方法.在提出的方法中,环同期预测法在时间序列数据基础上对月度定量预测;次级平台加权平均法完成对月度数据的修正,降低误差;数据留存评估法通过对数据量-有效数据量关系生成每月有效数据预测量;汇总成周期内有效数据预测量.理论分析及验证表明:多方法联合预测提供了"四川科技兴村"平台有效数据量的预测方法.
Abstract
The valid data of SiCHuan-TechRural-Revitalization-Plat relies on the moving average method,which cannot accurate-ly capture the real-time dynamic changes of data in practical applications,resulting in the prediction results not meeting the ex-pected accuracy and effectiveness.In order to effectively forecast valid data of SiCHuan-TechRural-Revitalization-Plat,a multi-method is proposed,which includes a contemporaneous prediction method,secondary platform weighted average method and data retention evaluation method.In the multi-method,the contemporaneous prediction method quantitatively predicts monthly data based on time series data;the secondary platform weighted average method corrects monthly data and reduces errors;the data re-tention evaluation method generates monthly effective data premeasurements by examining the relationship between data and valid data;summarize into valid data pre measurements within one cycle.Theoretical analysis and experimental results demonstrate that this provides an effective method for forecasting the valid data of SiCHuan-TechRural-Revitalization-Plat.
关键词
环同期预测法/次级平台加权平均法/数据留存评估法/移动平均法/多方法联合预测/有效数据量Key words
contemporaneous prediction method/secondary platform weighted average method/data retention evaluation meth-od/moving average method/forecasting of multi-method Integration/valid data引用本文复制引用
出版年
2024
NETL
NSTL
万方数据