Physics-constrained indirect supervised learning
This addresses the challenge of label scarcity in supervised learning for domain-specific applications like geophysics, though it appears incremental as it builds on existing indirect learning concepts.
The authors tackled the problem of supervised learning without direct labels by proposing a physics-constrained indirect supervised learning method that uses variables associated with labels as indirect labels and trains models with a physics-constrained loss, verifying its effectiveness on a well log generation problem.
This study proposes a supervised learning method that does not rely on labels. We use variables associated with the label as indirect labels, and construct an indirect physics-constrained loss based on the physical mechanism to train the model. In the training process, the model prediction is mapped to the space of value that conforms to the physical mechanism through the projection matrix, and then the model is trained based on the indirect labels. The final prediction result of the model conforms to the physical mechanism between indirect label and label, and also meets the constraints of the indirect label. The present study also develops projection matrix normalization and prediction covariance analysis to ensure that the model can be fully trained. Finally, the effect of the physics-constrained indirect supervised learning is verified based on a well log generation problem.