LASSO-based diagnosis scheme for multistage processes with binary data

•We study the fault diagnosis problem in multistage processes with categorical responses.•We start from a binary state-space model and frame the fault diagnosis as a variable (model) selection problem.•A practical LASSO-based diagnostic procedure which combines an extension of BIC with the adaptive LASSO algorithm is proposed.•We use a real example to demonstrate how the proposed procedure is used.

Process monitoring using multistage processes with binary data remains an important and challenging problem in statistical process control (SPC). Although the multistage processes has been extensively studied in the literature, the challenges associated with designing diagnostic schemes when only binary responses are observed are yet to be addressed well. This paper develops a practical LASSO-based diagnostic procedure which combines BIC with the popular adaptive LASSO variable selection method. Given the oracle property of LASSO and its algorithm, the diagnostic result can be obtained easily and quickly. More importantly, the proposed method does not require making any extra tests that are necessary in existing diagnosis methods. Numerical and real-data examples demonstrate the effectiveness of our method.