Investigation of Transfer Learning for Tunnel Support Design

The potential of machine learning (ML) tools for enhancing geotechnical analysis has been recognized by several researchers. However, obtaining a sufficiently large digital dataset is major technical challenge. This paper investigates the use transfer learning, powerful ML technique, used overcoming size limitations. study examines two scenarios where applied to tunnel support analysis. first scenario transferring knowledge between ground formation that well-studied new with very limited data. second intended investigate whether possible from relies on simplified more complex and realistic process involves training an Artificial Neural Network (ANN) adding extra layer model. added then trained smaller datasets fine-tune demonstrates effectiveness both scenarios. On this basis, it argued that, further development refinement, could become valuable tool ML-related applications.