Unsupervised novelty pattern classification of shmoo plots for visualizing the test results of integrated circuits
作者:
Highlights:
• Shmoo plots in semiconductor manufacturing are used for indicating device state.
• We address high-dimensional and multiclass imbalance challenges for shmoo plots.
• We introduce a feature extraction process for presenting clear shmoo plot patterns.
• We propose a two-stage clustering process to solve multiclass imbalance situations.
• To demonstrate the applicability of the proposed model, real field data is used.
摘要
•Shmoo plots in semiconductor manufacturing are used for indicating device state.•We address high-dimensional and multiclass imbalance challenges for shmoo plots.•We introduce a feature extraction process for presenting clear shmoo plot patterns.•We propose a two-stage clustering process to solve multiclass imbalance situations.•To demonstrate the applicability of the proposed model, real field data is used.
论文关键词:Shmoo plot,Class imbalance,High dimensionality,Unlabeled data,Feature extraction,Clustering
论文评审过程:Received 23 March 2021, Revised 14 March 2022, Accepted 25 April 2022, Available online 28 April 2022, Version of Record 2 May 2022.
论文官网地址:https://doi.org/10.1016/j.eswa.2022.117341