- Prediction candidate presentation
- Numerical analysis
Equipment log cross-cutting analysis technology for anomaly diagnosis and predictive detection
Pick up on data changes during operation and differences in items between bases, for use in quality control and predictions.
- We have developed a machine learning technology for analyzing tabular data recording multiple data items.
- Conventional tabular data analysis assumed that all data column configurations were the same.
- Using this newly developed machine learning technology, it is no longer necessary to rebuild models when data changes during operations or to build models for each base with differing data items. Predictions and judgments can be made using just one model.
Applications
- Predict failures based on equipment operation log
- Make defect judgments and predict status based on product inspection data
- Detect unauthorized access based on security log
Benchmarks, strengths, and track record
- Confirmed that the proposed method is able to categorize four types of public data sets*1 that prior research*2 failed to handle.
- Connect-4: Connect Four results category; splice: DNA sequence category; kr-vs-kp: Chess result prediction; PhishingWebsites: Phishing site detection
- Wang and Sun “TransTab: Learning Transferable Tabular Transformers Across Tables” NeurIPS 2022
Inquiries
Inquiries to Toshiba Corporate Laboratory (Komukai region)
Please include the title “Toshiba AI Technology Catalog: Equipment log cross-cutting analysis technology for anomaly diagnosis and predictive detection” 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.
