Semi-supervised Knowledge Transfer for Deep Learning from Private Training Data · HackerPapers discussions

Semi-supervised Knowledge Transfer for Deep Learning from Private Training Data — mentioned by obastani

Wed, 06 Dec 2017 21:33:53 GMT

Google has also done some foundational research on learning with privacy: https://arxiv.org/abs/1610.05755 https://research.google.com/pubs/pub45428.html It's hard to know for sure what they (or any company) are doing in production, but I'd be surprised if they don't use these things internally.

Paper and discussion history. A mention is not an endorsement.

Semi-supervised Knowledge Transfer for Deep Learning from Private Training Data — mentioned by haraldurt

Thu, 02 Nov 2017 23:38:28 GMT

In one sense, no. You can guarantee privacy of any given input (or any subset of k inputs) by applying transfer learning of an ensemble of models trained on subsets of the training data [0][1]. This is useful if, for instance, you train on medical data and you don't want anyone to know that "John Doe, HIV+" was part of the input. If your adversary does not take such precautions, however, then your canary should work. [0] https://arxiv.org/abs/1610.05755 [1] https://static.googleusercontent.com/media/research.google.c...

Paper and discussion history. A mention is not an endorsement.

Semi-supervised Knowledge Transfer for Deep Learning from Private Training Data — HN submission

Tue, 25 Oct 2016 12:49:36 GMT

Semi-Supervised Knowledge Transfer for Deep Learning from Private Training Data

80 points; 2 discussion replies on HN. Paper and source history. A submission is not an endorsement.