Towards Making Systems Forget with Machine Unlearning

2015 IEEE Symposium on Security and Privacy paper by Yinzhi Cao and Junfeng Yang that named machine unlearning: removing the effect of particular training data from a learned model without retraining it from scratch.