We have explored a few interesting ANN models in this chapter. These models can be applied to solve both supervised and unsupervised machine learning problems. The following are some of the other points that we covered:
- We have explored the necessity of ANNs and their broad types, that is, feed-forward and recurrent ANNs.
- We have studied the multilayer perceptron ANN and the backpropagation algorithm used to train this ANN. We've also provided a simple implementation of the backpropagation algorithm in Clojure using matrices and matrix operations.
- We have introduced the Enclog library that can be used to build ANNs. This library can be used to model both supervised and unsupervised machine learning problems.
- We have explored recurrent Elman neural networks, which can be used to produce ANNs with a small error in a relatively less number of iterations. We've also described how we can create and train such an ANN using the Enclog library.
- We introduced SOMs, which are neural networks that can be applied in the domain of unsupervised learning. We've also described how we can create and train an SOM using the Enclog library.
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.