John Hertz, Anders Krogh, Richard G. Palmer.
Introduction to the Theory of Neural Computation.
Addison Wesley. 1991
(read but not reviewed)
Contents:
- Introduction
- The Hopfield model
- associative memory • statistical mechanics • stochastic networks
- Extensions of the Hopfield model
- Optimization problems
- Simple perceptrons
- feed-forward networks • proof of convergence • linear, non-linear, stochastic units
- Multi-layer networks
- back-propagation
- Recurrent networks
- Boltzmann machines
- Unsupervised Hebbian learning
- principal component analysis • self-organizing feature extraction
- Unsupervised competitive learning
- adaptive resonance theory
- Formal statistical mecanics of neural networks