Sales Forecasting Using Convolution Neural Network

Sales forecasting is an essential component of business management, providing insight into future sales and revenue. It is critical for effective inventory management, cash flow, and business growth planning. While many retailers rely on simple Excel functions or subjective guesses from management, the industry is increasingly turning to machine learning techniques to develop more accurate and reliable prediction models. Among these techniques, Convolutional Neural Networks (CNN) emerged as a suitable option due to their ability to learn and improve accuracy over time. CNN applies several layers to make predictions, adjusting their weights with each input data point to minimize prediction error. As a result, sales forecasting with neural networks can significantly improve market operations and productivity for businesses. The validity of the proposed model is compared with the Facebook Prophet method, which is known as the recent time series forecasting method.