- Numerical analysis
- Causal Inference
Defect factor analysis using Transfer Lasso
Reduces the time and effort required for regular factor analyses, and identifies the essential factors of defects.
- In periodic analysis of defective factors, utilizing the results of a previous analysis eliminates unnecessary repeated inspections.
- Reduces the effects of nonessential factor changes (noise).
- Used in Toshiba’s semiconductor manufacturing facilities.
Applications
- Defect factor analysis in semiconductor factories
- Defect factor analysis in other plants
Benchmarks, strengths, and track record
- Dramatically reduces the time required for detailed investigations of analysis results at semiconductor factories and other plants.
Inquiries
Please include the title “Toshiba AI Technology Catalog: Defect factor analysis using Transfer Lasso” or the URL in the inquiry text.
Please note that because this technology is currently the subject of R&D activities, immediate responses to inquiries may not be possible.
References:
- Developed defect factor analysis AI that reflects and learns the knowledge of front-line plant engineers (in Japanese)
- Takada, M., & Fujisawa, H. (2020). “Transfer Learning via L1 Regularization”. NeurIPS2020.
- Explanatory article (“Sexy Technology”) in Nikkei Robotics February 2021 edition