Demand forecasting with color parameter in retail apparel industry using artificial neural networks (ANN) and support vector machines (SVM) methods

In this study, product variety has been taken into account and sales forecasting has been performed by using artificial intelligence to minimize error rate, in the retail garment industry. In this context, artificial intelligence models such as artificial neural networks (ANN) and support vector machines (SVM) have been established and inferences from the datasets have been made. During the establishment of the models, datasets have been prepared with and without color details of the products, for nine different products as separately and one combined dataset which consists all products, then the forecast process was carried out. Thus 20 different models were established and compared. Along with color detail, other variables that may have an effect on the sales performance such as weather, gender, special days etc., have been added to proposed models. In the comparison of methods root mean square error has been taken into consideration. As a result of this study, it has been determined that ANN outperformed SVM on seven datasets out of ten for the datasets without color and their performances were even for the datasets with color. The reliability of this study has been increased by comparing the results of the methods.