The prediction model of Chinese labor demand change is based on generative adversarial deep learning (GADL) and multi-modal data fusion

With the rapid development of artificial intelligence (AI) technology and the upgrading of industrial structure, the demand changes in the labor market show more complex dynamic characteristics, and it is difficult for traditional statistical and machine learning methods to accurately predict market trends. In order to improve the prediction accuracy and model adaptability, this paper proposes a labor demand prediction model based on generative adversarial deep learning (GADL) and multi-modal data fusion. The model uses generative adversarial network (GAN) to generate supplementary data to compensate for data scarcity, combined with convolutional neural network (CNN) and Transformer to conduct multimodal feature extraction, and finally time series modeling by long and short-term memory network (LSTM). The experimental results show that the GADL model is better than the LSTM and ARIMA method in several key performance indicators, and the mean square error (MSE) is reduced to 5% of the traditional method, and the determination coefficient (R²) reaches 0.9966, which significantly improves the prediction accuracy. The research shows that GADL model can more effectively capture the changing trend of labor market and provide data support for the optimization of government employment policy, enterprise human resource planning and talent training in universities. The study in this paper provides a new methodology for labor market prediction and lays the foundation for subsequent research.