| Summary: | Using Python language and combined with data analysis and mining technology, the authors capture and clean the housing source data of second-hand houses in Chengdu from Beike Network, and visually analyse the cleaned data. Then, a Random Forest (RF) model is established for 38,363 data elements. According to the visual analysis results, the model variables are revalued, the key factors affecting house prices are studied and the optimised model is used to predict house prices. The experiment shows that the deviation between the house price predicted by the RF model and that predicted by the real house price is small; it also indicates the accuracy of the RF model and demonstrates its good application value. © 2022 Yan Zhang et al., published by Sciendo.
|
|---|
