Neural Networks: What are "good" representations?

When training Neural Networks (e.g. FFNNs, Restricted Boltzmann Machines), lots of regularizers and other tricks can be applied to improve the results. Such tricks as well as different training...

02 February 2016 8,930 3 View