Physics Magazine @physicsmagazine.aps.org · Dec 18 Data-driven approaches to machine learning generally yield black boxes whose deductions are inaccessible. To address that shortcoming, researchers have designed a machine-learning method that can discover simple, human-interpretable laws from data. physics.aps.org/articles/v18...
5 likes 1 replies
?
Replies
Surely, though, the magic is the fusion, the multi-faceted connections (non-linear) of information to obtain a result, as with the brain. This would essentially consider the entire brain a 'black box' if it's a lack of clarity as to the how's and whys of retrieval.