然而，此类模型缺乏从消费者个体角度分析新产品成功扩散的原因，如消费者个

However, such models lack the analysis of the reasons for the successful spread of new products from the perspective of individual consumers, such as individual consumers' different preferences for new products, and differences in social networks where consumers are located. The heterogeneous subject model (HAM) can overcome the above limitations [5], and can calculate the purchase probability of individual consumers considering the way of releasing new product information. At the same time, the empirical estimation of the parameters of HAM based on the real data of movie products overcomes the defect of "toy model". Therefore, this paper integrates the improved infectious disease model with HAM to obtain a new product sales forecast coupling model.

However, such models lack to analyze the reasons for the successful spread of new products from the individual perspective of consumers, such as the different preferences of individual consumers for new products, the differences in the social networks in which consumers live, and so on. The heterogeneous subject model (HAM) can overcome the above limitations and calculate the probability of individual consumers buying when considering the way new product information is released. At the same time, the "toy model" is overcome by empirical estimation of HAM parameters by the real data of film products. Therefore, this paper integrates the improved infectious disease model with HAM to obtain a new product sales prediction coupling model.

However, such models lack the analysis of the reasons for the successful diffusion of new products from the perspective of individual consumers, such as consumers' different preferences for new products, and the differences of consumers' social networks. The heteromorphic agent model (HAM) can overcome the above limitations [5], and can calculate the purchase probability of individual consumers considering the new product information release mode. At the same time, through the real data of film products, the parameters of ham are estimated empirically, which overcomes the defect of "toy model". Therefore, this paper integrates the improved epidemic model and ham to get a new coupling model of product sales forecasting.<br>