Unlearning in Tabular-to-Hypergraph Learning via Selective Distillation

Machine unlearning requires efficiently removing the influence of specified information (rows/columns/values) from a deployed model to satisfy privacy and regulatory constraints. Converting tables into hypergraphs can capture high-order relations such as “same column”, “same value”, and multi-column interactions, but unlearning for hypergraph-based models is often implemented via costly re-training or gradient/Hessian-based approximations, and becomes especially chal-lenging for column- and value-level deletion due to broad structural dependencies and limited verifiability. We propose HERMES, a fundamentally different un-learning paradigm for tabular-to-hypergraph modeling based on selective knowledge transfer rather than parameter rollback or second-order correction. HERMES freezes the original trained model as a teacher and trains a student as the released unlearned model: on the retained set, the student learns from the teacher through structure-augmented distillation and consistency regularization to pre-serve utility; on the forget set, the student is driven toward maximum-entropy pre-dictions and explicitly pushed away from the teacher via anti-distillation diver-gence, activ