Abstract: This study explores alternative framework configurations for adapting thermal machine learning (ML) models for power converters by combining transfer learning (TL) and federated learning (FL ...