A neural network approach for predicting manufacturing performance using knowledge management metrics

Abstract

This paper aims to devise a model for predicting the knowledge management (KM) effect on manufacturing performance via neural network (NN). This is the first empirical study that applies NN to forecast manufacturing performance using 48 KM metrics which cover knowledge resources, KM processes, and KM factors. The training, validation, and testing of the NN model were based on 580 usable data points of KM and manufacturing performance collected from manufacturing companies in Malaysia. The research findings reveal that the NN model serves as a reliable yet simple tool to predict the manufacturing performance of a company by considering various essential KM metrics. The network prediction is in good correlation with the actual data. Lastly, the prediction model will be useful for practitioners to determine future KM strategies and targets to improve manufacturing performance.