Simulation of cigarette sales forecasting model based on deep learning and big data algorithm

As deep learning technology and big data analytics continue to advance, sales forecasting has become increasingly important in the business sector. This paper is devoted to the application and simulation of cigarette sales forecasting model based on deep learning and big data algorithm. In this study, taking cigarette sales as a case, a convolutional neural network (CNN) model utilizing the Attention mechanism (AE) was developed to analyze historical sales data on a large scale. This model was designed to identify sales trends and patterns by integrating known information about the forecasted time point, offering a new perspective on sales forecasting, such as whether the time point is a holiday or not. In this paper, the known information of predicted time points is fused, and it is spliced with the results of AE to get the input of decoder stage. Through the simulation experiment of the model, we verify the high accuracy of deep learning and the information richness of big data analysis. The findings indicate that the model is capable of accurately forecasting short-term sales patterns, but also make long-term sales forecasts, which is helpful for enterprises to better formulate sales strategies and resource planning. This research strongly supports the use of deep learning and big data algorithms for sales forecasting, offering valuable insights for both future studies and practical applications.