Ensemble convolutional neural networks with weighted majority for wafer bin map pattern classification
- PDF / 1,603,161 Bytes
- 14 Pages / 595.276 x 790.866 pts Page_size
- 85 Downloads / 313 Views
Ensemble convolutional neural networks with weighted majority for wafer bin map pattern classification Chia-Yu Hsu1
· Ju-Chien Chien2,3
Received: 9 February 2020 / Accepted: 1 October 2020 © Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract Wafer bin maps (WBM) provides crucial information regarding process abnormalities and facilitate the diagnosis of low-yield problems in semiconductor manufacturing. Most studies of WBM classification and analysis apply a statistical-based method or machine learning method operating on raw wafer data and extracted features. With increasing WBM pattern diversity and complexity, the useful features for effective WBM recognition are highly dependent on domain knowledge. This study proposes an ensemble convolutional neural network (ECNN) framework for WBM pattern classification, in which a weighted majority function is adopted to select higher weights for the base classifiers that have higher predictive performance. An industrial WBM dataset (namely, WM-811K) from a wafer fabrication process was used to demonstrate the effectiveness of the proposed ECNN framework. The proposed ECNN has superior performance in terms of precision, recall, F1 and other conventional machine learning classifiers such as linear regression, random forest, gradient boosting machine, and artificial neural network. The experimental results show that the proposed ECNN framework is able to identify common WBM defect patterns effectively. Keywords Wafer bin map · Deep learning · Convolutional neural network · Ensemble classification · Weighted majority · Semiconductor manufacturing
Introduction With the rapid development of semiconductor manufacturing technology, controlling the production process effectively is critical for minimizing process variation to enhance yield (Chien et al. 2013; Hsu 2014). Circuit probe (CP) testing is used to evaluate each die on the wafer after the wafer fabrication processes. Wafer bin maps (WBMs) represent the results of a CP test and provide crucial information regarding process abnormalities, facilitating the diagnosis of low-yield problems in semiconductor manufacturing (Hsu and Chien 2007; Chien et al. 2013; Hsu 2015). A WBM is a two-dimensional
B
Chia-Yu Hsu [email protected]
1
Department of Industrial Engineering and Management, National Taipei University of Technology, Taipei, Taiwan
2
Department of Computer Science, National Tsing Hua University, Hsinchu 30013, Taiwan
3
Artificial Intelligence for Intelligent Manufacturing Systems (AIMS) Research Center, Ministry of Science & Technology, Hsinchu 30013, Taiwan
defect pattern which is transformed into binary values and used to select the testing bin code. The dies that pass the functional test are denoted as 0 and the defective dies are denoted as 1. Depending on the various sources of variation, the WBM consists of random, systematic, or mixed defects generated during semiconductor fabrication (Hsu and Chien 2007; Hsu et al. 2020). Random defect patterns are caused by random pa
Data Loading...
