Yield Prediction Model of Spinneret Buds Based on Stacking Algorithm

This paper proposes to integrate single models such as random forest (RF), BP neural network (BPNN), and support vector regression (SVR) with the integration algorithm of Stacking and average, and constructs four kinds of integration models: The RF-BPNN-Stacking model is constructed using the stochastic forest and BP neural network models and the Stacking integration algorithm. The RF-BPNN-SVR-Stacking model is constructed by Stacking integration algorithm based on random forest, BP neural network and support vector regression model. The RF-BPNN-AVG model is constructed by integrating average method with random forest and BP neural network model. The RF-BPNN-SVR-AVG model is constructed by means of average integration algorithm with random forest, BP neural network and support vector regression model. The prediction effect of four integrated models is compared with that of a single model. The experimental results show that the prediction effect of the four integrated models constructed in this paper is better than that of the single model, and the prediction accuracy can be improved when applied to the yield of spina bud.