An Introduction to Machine Unlearning
This work provides a foundational overview and standardization for the emerging field of machine unlearning, which is crucial for researchers and practitioners dealing with data removal challenges in ML systems.
The paper tackles the problem of efficiently removing the influence of specific training data from machine learning models to address privacy, fairness, and data quality issues, by summarizing and comparing seven state-of-the-art machine unlearning algorithms, consolidating definitions, reconciling evaluation approaches, and discussing practical applications.
Removing the influence of a specified subset of training data from a machine learning model may be required to address issues such as privacy, fairness, and data quality. Retraining the model from scratch on the remaining data after removal of the subset is an effective but often infeasible option, due to its computational expense. The past few years have therefore seen several novel approaches towards efficient removal, forming the field of "machine unlearning", however, many aspects of the literature published thus far are disparate and lack consensus. In this paper, we summarise and compare seven state-of-the-art machine unlearning algorithms, consolidate definitions of core concepts used in the field, reconcile different approaches for evaluating algorithms, and discuss issues related to applying machine unlearning in practice.