Understanding deep learning requires re-thinking generalization

Re:Understanding deep learning requires re-thinking generali... by RichardHeart on Saturday May 13, 2017 @03:35AM (#14329925)

Would it be possible to use machine learning to do this job better, meaning, could the machines look at what other machines are doing, and better translate to us whats going on?

[ Parent | on HN ]

Re:Understanding deep learning requires re-thinking generali... by fooker on Saturday May 13, 2017 @05:30AM (#14330124)

Use a system we don't understand to understand another system we don't understand; what could go wrong?

[ Parent | on HN ]

Re:Understanding deep learning requires re-thinking generali... by jacquesm on Saturday May 13, 2017 @05:44AM (#14330150)

> what could go wrong?

That's already known: there are inputs to the network that do not make sense and yet will trigger strong responses. Think of them as inputs that have the same effect on NNs that optical illusions have on the human brain. We infer something that isn't there.

I suspect that as network architectures get better and parameter counts drop these will get harder and harder to construct.

[ Parent | on HN ]

Re:Understanding deep learning requires re-thinking generali... by Franciscouzo on Saturday May 13, 2017 @11:09AM (#14331347)

You could use a neural network to optimize the hyperparameters of another neural network, but afaik it's not a really efficient way, your best bet would be to use bayesian hyperparameter optimization.

[ Parent | on HN ]