Birth Rate Prediction System Based on Grey Prediction and Gradient Boosting Regression Tree Ensemble Model

Abstract

We selected several factors affecting the Birth Rate in Guangdong Province. Then we saw the contribution of each characteristic to the Birth Rate by the gradient boosting regression tree model. The grey prediction algorithm used each characteristic to predict its value in a certain year. Using variance ratio and small residual probability, we evaluated the prediction accuracy. We used the gradient boosting regression tree for predicting the Birth Rate in the next two years. This was done under the premise of knowing the data of each characteristic. The thinking of the factors affecting the Birth Rate and the prediction of the Birth Rate prompt us to think about the relevant realistic factors. They also assist the government in focusing on the adjustment of fertility policy. The aim is to promote population development and social progress to the greatest extent possible.