Layers and hidden representations
A neural network does not separate classes. It bends the space until the classes become separable. This whole lesson is about showing that bending.
1 steps
55 XP
Open without an account
Start the lesson →
Sources
- Olah, C. 2014 · Neural Networks, Manifolds, and Topology · colah.github.io
- Hornik, Stinchcombe, White 1989 · Multilayer Feedforward Networks are Universal Approximators · Neural Networks, 2(5)
- Goodfellow, Bengio, Courville 2016 · Deep Learning, Bölüm 6 · MIT Press
ML Academy · an interactive machine learning course that runs in your browser ·
All lessons