Study on prediction models for integrated scheduling in semiconductor manufacturing lines

Prediction of machine breakdowns and lot product qualities plays an important role in scheduling of semiconductor manufacturing lines because accurate state of machines and lot are a key prerequisite for good scheduling. The difficulties in modeling result from the datum with high dimensions and stochastic noise that are brought inevitably in actual environment. This paper presents a novel prediction model referred as Incremental Extreme Least Square Support Vector Machine (IELSSVM), which transforms the data into Extreme Learning Machine (ELM) space and then minimize the structural risk like LSSVM. The transformation into ELM feature space can be regarded as a good dimensionality reduction. The incremental formula is proposed for on-line industrial application to avoid retraining when data comes one by one or chunk by chunk.