Research on the Application of PCA-BPNN Algorithm in Housing Price Prediction
Housing prices is an important factor affecting people's happiness index,so it is of great signifi-cance to predict housing prices reasonably.Taking the classic prediction datasets-the Boston House Price Datasets-as an example,an improved algorithm PCA BPNN based on principal component analysis(PCA) for a 3-layer BP neural network model is proposed for house price prediction.On the basis of data standard-ization and principal component analysis dimensionality reduction on the datasets,the prediction model is optimized by adjusting parameters such as the number of hidden layer neurons and learning times of the BP neural network model.Finally,it uses MATLAB to conduct simulation experiments on the data.The ex-perimental results show that the proposed model has improved prediction accuracy compared to the original BP neural network model,with a maximum improvement of 90.4772%.