부산대학교 도서관

The price prediction of building materials is an important issue for the construction industry and related industries. At the same time, the price fluctuation of building materials has an important impact on the result of the project cost. CatBoost algorithm is a gradient lifting tree algorithm, with high efficiency and accuracy, suitable for various types of data. In this paper, CatBoost algorithm is used to establish the price prediction model of building materials. By collecting the data of Φ16‐25mm, HRB400 type rebar construction material price and related influencing factors, data cleaning and feature engineering are carried out, the data is divided into training set and test set, and CatBoost algorithm is used to train and optimize the model. The experimental results show that CatBoost algorithm has high accuracy and efficiency in predicting the price of building materials.