Prediction of Cloth Waste Using Machine Learning Methods in the Textile Industry

The modern methods offered by the field of artificial intelligence bring an alternative solution to the problems existing in the industry. In this study, it is aimed to make an estimation of cloth waste in a textile company by using machine learning models. Thus, instead of producing with a fixed waste rate, the company evaluates many different parameters affecting the production process and obtains waste rates that vary according to the order. In this way, the company can overcome the problems that may be encountered in terms of both resource management and cost with the right amount of production. In the study, four machine learning methods (Support Vector Regressor, Random Forest, Decision Tree Regressor, and Bagging) were used and the results show that bagging gives the highest R2 value and minimum prediction error with 0.86 R2 when compared to other approaches.