Smart home systems intended for end-users will require a way of conveying why they selected a specific (past or future) action. This will build trust in such autonomous distributed systems because inhabitants of smart homes should be able to find out why their devices are behaving in a certain way. A similar challenge has been identified for machine learning systems in general and researchers are in particular targeting the explainability of deep learning systems The aim of this master thesis is to apply the existing techniques for explainability to the smart home domain.

Please contact me for more information on this topic.