Multi-layer perceptron (machine learning)
Dear team, I recently developed an interest in network analysis. Most of the network psychometric approaches I have read about rely war on single-layer networks. A colleague recently mentioned that a multi-layer perceptron present alternative to the clustering (community detection) usually used with networks. It seems that this method can be implemented using this function:
R: Create and train a multi-layer perceptron (MLP) (r-project.org)
Do you think this usage is appropriate? Is it available in JASP? I am very much nought the expert on neural networks, so I'm assuming that neural network regression is related, but in some way distinct? Apologies, if this is just another name for the same thing!
Many thanks in advance!
Comments
JASP's machine learning module supports networks with arbitrary numbers of hidden layers. (Note: A "perceptron" is a historical example of a neural net.)
R