An artificial neural network-based model was used for predicting components required in the manufacture of hot melt adhesives. Use of many ethylene-vinyl acetate copolymers and hydrocarbon resins was taken into consideration to optimize the comps. and replace the new components with those available at company storage.