Prediction Method for Domestic Used Car Selling Prices Based on CNN-BiLSTM

To address the issue of low accuracy in predicting used car prices, a prediction model combining Convolutional Neural Network (CNN) and Bidirectional Long Short-Term Memory (BiLSTM) is proposed. The model takes standardized used car data as input and establishes a mapping between input and output by leveraging CNN (1-dimensional) for feature extraction and BiLSTM for time-series processing in the composite model. Furthermore, Random Forest and Support Vector Regression (SVR) are selected as comparative models to evaluate the predictive capability of the CNN-BiLSTM model. Experimental results demonstrate that the CNN-BiLSTM model achieves the highest accuracy, with an accuracy rate of 0.923, indicating the superior predictive performance of the model. This study can serve as a reference for practical used car price assessment.