Stacking strategy-assisted random forest algorithm and its application

Short-term power load forecasting provides important guidance for the improvement of marketing and control levels enterprises. In this paper, a novel method, named RF-TStacking, is proposed to forecast short-term load. This study starts from influence factors load, random forest applied estimate importance Based on Stacking strategy, integration LightGBM realized achieve forecasting. To improve generalization ability model, put back sampling used sample each primary learner, average value taken as result learner. The Bayesian optimization adjust super parameters model accuracy selection influencing factors. data region in northwest China are testing, it found that can provide stable prediction results.