What is the difference between recurrent neural networks and Hopfield networks, or are they same?
Best Answer
They are not the same. A Hopfield network is one particular type of recurrent neural network.
Take a look at Chapters 14 and 15 of Haykin, Neural Networks. A recurrent neural network is any neural network in which neurons can be connected to other neurons so as to form one or more feedback loops (i.e. not like in a multilayer perceptron where everything goes one way – see the pictures in this question.) There are basically two useful kinds of recurrent network at the moment. One kind are those that try to simulate the human memory. The Hopfield network is a particular example. The other kind are input-output mapping networks, which can be used for calssification and prediction of time series data.
Similar Posts:
- Solved – Differences between Multi-layer NN, Hopfield, Helmholtz and Boltzmann machines
- Solved – Differences between Multi-layer NN, Hopfield, Helmholtz and Boltzmann machines
- Solved – Difference between SOM and Hopfield
- Solved – Self Learning Neural network
- Solved – the relation between belief networks and Bayesian networks