Knowledge extraction from deep convolutional neural networks applied to cyclo-stationary time-series classification

Mathematics

• Knowledge Extraction 100%
• Stationary Time Series 69%
• Condition-based Maintenance 51%
• Neural Networks 51%
• Learning 49%
• Time series 48%
• Gearbox 47%
• Fault Diagnosis 42%
• Knowledge 41%
• Methodology 40%
• Fault 31%
• Architecture 30%
• Training 29%
• Necessity 28%
• Data Structures 27%
• Configuration 21%
• Necessary 18%
• Model 17%
• Modeling 17%
• Performance 16%

Business & Economics

• Knowledge Extraction 83%
• Stationary Time Series 82%
• Deep Learning 75%
• Neural Networks 65%
• Fault Diagnosis 46%
• Condition-based Maintenance 43%
• Learning Model 41%
• Data Structures 34%
• Fault 31%
• Methodology 27%
• Performance Improvement 27%
• Severity 26%
• Modeling 16%

Engineering & Materials Science

• Deep neural networks 53%
• Convolutional neural networks 49%
• Deep learning 48%
• Time series 47%
• Failure analysis 23%
• Data structures 22%