There are applications that may require removing the trace of a sample from system, e.g., user requests their data to be deleted, or corrupted is discovered. Simply storage units does not necessarily remove its entire since downstream machine learning models store some information about samples used train them. A can perfectly unlearned if we retrain all it scratch with removed training dataset...

The iterative unlearning algorithm for connectivity self~correction is proposent. No

Article history: Received 18 January 2013 Received in revised form 1 July 2013 Accepted 13 December 2013 Available online 24 January 2014 Radical innovations have enormous impacts on organizations, industries, and societies. The success of a radical innovation requires multiple facilitators within and across organizational boundaries. In this study, we distinguished the effects of two dimension...

144 (short abstract) and 218 (long abstract) SHORT ABSTRACT In this work we reevaluate and elaborate Crick-Mitchison's proposal that REM-sleep corresponds to a self-organized process for unlearning attractors in neural networks. This reformulation is made at the face of recent findings concerning the intense activation of the amygdalar complex during REM-sleep, the involvement of endocannabinoi...

The right to erasure requires removal of a user's information from data held by organizations, with rigorous interpretations extending downstream products such as learned models. Retraining scratch the particular omitted fully removes its influence on resulting model, but comes high computational cost. Machine "unlearning" mitigates cost incurred full retraining: instead, models are updated inc...

The leave-one-out cross-validation scheme for generalization assessment of neural network models is computationally expensive due to replicated training sessions. In this paper we suggest linear unlearning of examples as an approach to approximative cross-validation. Further, we discuss the possibility of exploiting the ensemble of networks o ered by leave-one-out for performing ensemble predic...