Multi-source domain adaptation via evidence-based target pseudo-labels for human–computer collaboration fault diagnosis

Key Points

• Enhanced fault diagnosis leverages target pseudo-labels to increase accuracy.
• The model showed a significant improvement, with accuracy rising by 20% in tests.
• Analysis utilizes multi-source domain adaptation to integrate various training data effectively.
• This approach highlights the potential for better collaboration between humans and computers in diagnostics.