Tom Breloff

Agreed. This needs cleanup. I like your idea to do all the testing here! (I really need to improve VRT) On Wednesday, August 24, 2016, Christof Stocker [email protected] wrote: >...

Additionally, what would be really powerful is if we didn't need to subtype from Transformation at all, as long as the query functions were implemented for the type, and anything...

> A way to dispatch on a bias being present. Maybe name the types Linear... vs Affine... If the type is parameterized, then `typealias Linear{W} Affine{W,Void}`? > for performance reasons...

> T := y = N(wx,sigma) I like this, if we can figure out the right way to do it. It should probably be `T := y = N(wx +...

Yeah I guess we should have our convention be `y = x'w + b`, which is common in Matlab workflows

I pushed up some of my experiments... I think they're pretty promising. Check out the main [Transformations.jl in the tom branch](https://github.com/JuliaML/Transformations.jl/tree/tom/src/Transformations.jl). The gist is that there's a `immutable Transformation` which...

Hi Tim. Yes computational graph is part of it. I do care about backprop, though not everyone will. It's more an attempt to generalize and unify different models and approaches....

For reference, copying from gitter: > Tom Breloff @tbreloff > i pushed up a mostly-working prototype to Transformations... i'd love comments/criticism on the approach so far > if we have...

One piece I'm a little unsure about is how best to add penalties to the mix. Are the penalties a component in each learnable transformation? Or maybe they are passed...

I'm going to move on to ObjectiveFunctions next week, and try to piece some of this stuff together. I'm thinking that package will be no more than conveniences to connect...