大规模数据环境下用电量预测算法研究

The Research of Electricity Using Prediction Algorithm Based On Intelligent Electricity Network

提出了在海量数据下的分布式局部线性回归算法,并通过在大规模用电记录数据集上进行详细地实验,证明了该算法在保证预测效果的前提下,能有显著地缩短模型训练时间,并具有很好的可扩展性。

This paper proposes a distributed local linear regression algorithm in the mass data,and by in large electric log data set detailed experiments,proves that the algorithm ensures prediction effect,can significantly shorten the training time model,and has good scalability.