Comparison Analysis of Facebook’s Prophet, Amazon’s DeepAR+ and CNN-QR Algorithms for Successful Real-World Sales Forecasting

By successfully solving the problem of forecasting, processes in work various companies are optimized and savings achieved. In this process, analysis time series data is particular importance. Since creation Facebook's Prophet, Amazon's DeepAR+ CNN-QR forecasting models, algorithms have attracted a great deal attention. The paper presents application comparison above for sales distribution companies. A detailed performance over real with different lengths history was made. results show that Prophet gives better items longer frequent sales, while superiority without long rarely sold.