Rubber Spare Parts Supplier Selection Model Using Artificial Neural Network: Multi-Layer Perceptron

Abstract

Supplier have an important role in the availability of raw materials for the ongoing production activities of a company. The selection of the right supplier is not only profitable for the company but also can increase customer satisfaction. Therefore, in the selection of suppliers the company must have a system of selection and evaluation of suppliers of raw materials and components. The main purpose of the supplier selection process is to reduce purchasing risk, maximize overall value for buyers, and build long-term relationships between buyers and suppliers. Supplier selection model is used to facilitate the strategic direction of supply chain management to take several criteria from suppliers to achieve the priorities desired by the company. This research was conducted at a manufacturing company engaged in auto parts. In this study the problem is that the company chooses suppliers to supply raw materials based solely on the list of suppliers who are willing to agree on the price offered by the company with suppliers so that the company has difficulty choosing suppliers to become long-term suppliers and delays in the supply of raw materials from each -one supplier. This study aims to create a supplier selection model framework to classify each supplier so that it can be used as a supplier in the long term. This study uses the Artificial Neural Network (ANN) method with the multi layer perceptron classification technique. Artificial Neural Network is used to create an efficient and inefficient supplier selection model. The accuracy of the ANN model is 85.9756%, the statistical kappa value is 0.7152, with an MAE error value of 0.1478, the MSE error value is 0.3107. From the ANN obtained 4 criteria used in supplier selection, namely the criteria of quality, delivery, price, and warranty and complaint services.