Survey of Machine Learning and Deep Learning Approaches on Sales Forecasting

Abstract

Sales forecasting plays an important role in the modern financial system. It is used in the private and government financial institutions, companies, industries, factories, trading, etc. Due to the necessity of sales forecasting, this study focused the machine learning and deep learning approaches used for predicting the future sales and demands. These approaches accept the input as historical sales data and generate the response as future demands. From this survey it is observed that deep learning approaches are performed better than machine learning approaches in terms of prediction accuracy. In deep learning approaches, Convolutional Neural Network (CNN) can attain high prediction accuracy