Predicting job completion time in a wafer fab with a recurrent hybrid neural network

Predicting the completion time of a job is a critical task to a wafer fabrication plant (wafer fab). Many recent studies have shown that pre-classifying a job before predicting the completion time was beneficial to prediction accuracy. However, most classification approaches applied in this field could not absolutely classify jobs. Besides, whether the pre-classification approach combined with the subsequent prediction approach was suitable for the data was questionable. For tackling these problems, a recurrent hybrid neural network is proposed in this study, in which a job is pre-classified into one category with the k-means (kM) classifier, and then the back propagation network (BPN) tailored to the category is applied to predict the completion time of the job. After that, the prediction error is fed back to the kM classifier to adjust the classification result, and then the completion time of the job is predicted again. After some replications, the prediction accuracy of the hybrid kM-BPN system will be significantly improved.