A hybrid fuzzy and neural approach for forecasting the book-to-bill ratio in the semiconductor manufacturing industry

Abstract

Accurately forecasting the book-to-bill (BB) ratio is critical to the semiconductor manufacturing industry. Therefore, a hybrid fuzzy linear regression and back-propagation network (BPN) approach is proposed in this study. In the proposed approach, multiple experts construct their own fuzzy multiple linear regression models from various viewpoints to forecast the future BB ratio. Each fuzzy multiple linear regression model can be converted into two equivalent nonlinear programming problems to be solved. To aggregate these fuzzy BB ratio forecasts, a two-step aggregation mechanism is applied. At the first step, a fuzzy intersection is applied to aggregate the fuzzy BB ratio forecasts into a polygon-shaped fuzzy number, in order to improve the precision. After that, a back-propagation network is constructed to defuzzify the polygon-shaped fuzzy number and to generate a representative/crisp value, so as to enhance the accuracy. The practical case of equipment manufacturers with their headquarters in North America is used to evaluate the effectiveness of the proposed methodology. According to the experimental results, the proposed methodology improved both the precision and accuracy of the BB ratio forecasting by 21% and 81%, respectively.