Research on Stacking multiple models for optimizing university library book procurement forecasting
This paper presents a book procurement forecasting model based on Stacking ensemble of multiple models,aiming to enhance the accuracy and reliability of book procurement forecasts in universities.Traditional single prediction models struggle to effectively account for the various complex factors in university book procurement.By employing the Stacking method,a second-ary model is constructed to efficiently integrate predictions from different base models,and the best Stacking model is chosen through cross-validation to ensure model stability and generalization.Experimental results demonstrate that the Stacking ensemble of multiple models significantly improves the accuracy and robustness of book procurement forecasts in universities.This offers an effective decision-making tool for university book procurement management,with the potential to enhance resource allocation,re-duce unnecessary costs,and elevate the scientific rigor of management decisions.
NETL
NSTL
万方数据
