CSenn

Yes, we will be adding a discovery tool and an import/ export tool :)

Hi Everyone,

I am looking for feedback on a side project some friends and I have been working on.

First, do you see any value in being able to create, collaborate with access control, version, and diff data designs, and then output them to different formats? I have done some work in healthcare, energy, manufacturing, and e-commerce and in every one have seen json/xml schemas (or free text data specs) being passed around through emails, word docs, excel spreadsheets, hosted s3 files, power points, custom built websites, etc. I’ve also seen bugs when versions get confused between data partners or comments/changes get lost in email threads. It’s also not fun to read and maintain a 300-3,000 line data schema file.

Second, do you seen any value in being able to compose data designs? I personally think the idea of reusing data structures and composing data designs with a few clicks is interesting, but I’m curious to the practicality of it?

Also the json-ld graph output feels like it could be wrong? I am trying to avoid constraining properties using rdfs:domain as is common in RDF (since properties can be shared between designs and it creates a coupled export), however lib:children seems a little off?

Usually the number of nodes in the input and output layers don't affect things all that much. They are relatively set based on the problem. The number of nodes in the hidden layers, the number of hidden layers, and various other parameters such as cost and activation functions, are mostly what you use to tune performance.

The largest float in the output layer (while the graph is yellow) is the activation. The largest activation in the final layer is the network's "guess". The guess is the index of the last layer, which corresponds to a particular digit.

Hmm, that sounds like a promising idea...

In my experience spinners work great for async calls, but not so great for blocking synchronous computations. There is probably something that could be done though with a little bit of creativity though...

That's a good point. I tried some light optimizations using ShouldComponentUpdate in React, but I believe the bottleneck is a synchronous blocking call in the D3 rendering process. Larger networks write up 25,000 SVG elements, and I was not sure how to significantly improve rendering speed. Someone suggested Canvas with D3 could speed rendering up?