SplitNN is a distributed and private deep learning technique to train deep neural networks over multiple data sources without the need to share raw labelled data directly.

The post Split learning: Distributed deep learning method without sensitive data sharing appeared first on Analytics India Magazine.