search

autumnai / leaf

Open Machine Intelligence Framework for Hackers. (GPU/CPU)
Apache License 2.0
5.56k stars 271 forks source link

name confuse #94

Open b0c1 opened 8 years ago

b0c1 commented 8 years ago

Hi guys, I'm a little confused. I'm a beginner in neural networks but some docs/library use neuron/preceptron/etc (like http://synaptic.juancazala.com/ ) some using layer/solver (leaf). What's the difference? Can I write a feedforward network in leaf? (like http://synaptic.juancazala.com/#/paint-an-image )

alexandermorozov commented 8 years ago

It looks like perceptron in synaptic corresponds to one Sequental layer in leaf that itself consists from 3 layers: 1) Linear, 2) Sigmoid or ReLU, 3) another Linear.

Currently Leaf supports feedforward networks and AFAIK only them. Here is an example.

MichaelHirn commented 8 years ago

Supporting @alexandermorozov's answer:

Feedforward networks are built with a Sequential layer in Leaf. The Sequential layer can use any kind of layer, even other Sequential layers. The API in Leaf is very, very easy. I would say even easier than the synaptic one.

There is currently a PR for the Leaf Book #93 which has all the chapters by now, although there are still a lot of grammar, spelling mistakes - clean up is on the way. The Leaf Book will also have a Glossary, I will use your input of neurons/perceptrons/etc. If there are some other confusions or if you would like to clarify the confusion a bit more, I would be pleased about a comment at #93