A memetic PSO based KNN regression method for cycle time prediction in a wafer fab

In this paper, cycle time prediction of wafer lots is studied. A memetic algorithm called GSMPSO by combining the PSO with a Gaussian mutation operator and a Simulated Annealing (SA)-based local search operator is developed to weight the features for K Nearest Neighbors (KNN) regression. The GSMPSO-KNN regression method is used to predict the cycle time of wafer lots. The experiment result demonstrates that a more accurate result can be obtained by the proposed method compared with some other prediction methods. The critical factors affecting the cycle time of wafer lots can also be extracted by the proposed method.