Hi! I would expect the implementation of any neural network to be dictated by the particular mathematical/algorithmic description. I am not at all sure what description might have given rise to your code. Do you have any particular type of neural network in mind? Or any particular task to perform?
I also wonder why you consider the construction of the network graph (as above) to be an inherently stateful activity. And why you choose to have those construction functions operating on a single global variable. When constructing a map in Clojure, one uses functions that take one immutable map as argument and return another immutable map, the 'updated' version, as return value. In the same way, why not design functions that take one immutable network graph as an argument and return another? I should probably mention that, although I know a little about neural networks, I know next to nothing about Clojure! That may be obvious already :) I do mean to install it one of these days... --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en -~----------~----~----~----~------~----~------~--~---
