一种基于Stacking集成机器学习的城市房租预测模型OA

Effective and accurate prediction models for urban housing rental prices play an important role for the government to make relative policies.In this paper,we propose a stacking ensemble machine learning model to improve the efficiency and accuracy superior to the existed models.Based on the rental-price data of Chengdu during 2022 and 2023,we firstly show that the stacking model is the most accurate among six ordi-nary machine learning models and their stacking model,while the stacking model has the lowest time effi-ciency simultaneously.Then we construct a new stacking model by selecting the XGBoost and RF models as the first layer base learners,which greatly improves the time efficiency of the stacking model.Finally,empiri-cal analysis verifies the high accuracy and time efficiency of the new stacking model.