Job cycle time estimation in a wafer fabrication factory with a bi-directional classifying fuzzy-neural approach

摘要: Estimating the cycle time for every job in a factory is critical task. It was recently reported that classification noticeably enhanced accuracy of estimation. In pre-classifying approaches, whether pre-classification approach combined with subsequent estimation suitable data questionable. Conversely, difficulty classifying according to only error not various attributes problem post-classifying approaches. To tackle these problems, bi-directional fuzzy-neural proposed this study. methodology, jobs are pre-classified but also post-classified. The results and post-classification aggregated into suitability index each job. A then assigned category which its highest. radial basis function network constructed predict attributes. evaluate effectiveness practical example used According experimental results, methodology significantly better than those many existing