5 ms·
Neural Networks Virtual Study Group
- almost 18y agoSeems like a great idea. Would be cool to see more of these... Of course C# seems like a horrible choice for learning about this sort of stuff, but maybe that's just me.
- paulgb 18y agoYeah, these distributed courses are a great idea. There's a bunch of free textbooks online, especially for CS and math topics. I'd love to see online communities develop around the textbooks for like-minded people to form study groups.
- samson 18y agoI was thinking the same thing as well. It would be great to see more of these done in different languages for various topics.
- signa11 18y agohmm, i don't really think C# has anything to do with it, other than just providing a vehicle for implementation of fundamental ideas of NN theory (imho ofcourse). in my opinion, using mathematica (or some such) would have been a far saner choice, to get to the essence of the subject. also, a dosage of classical (pioneering ?) work as described in (parallel-distributed-processing vol-1&2) http://www.amazon.com/Parallel-Distributed-Processing-Vol-Foundations/dp/026268053X/ref=sr_1_1?ie=UTF8&s=books&qid=1232857316&sr=1-1 http://www.amazon.com/Parallel-Distributed-Processing-Vol-Fo..., would be just great.
- psyklic 18y agoIt yells danger to me when only one week of course materials is not "under construction."
- dustineichler 18y agoquestion: where else can i learn about neural networks and ai?
- signa11 18y agomy recommendations (for neural networks) would be as follows: - parallel distributed processing (PDP group vol 1 & 2) - simon haykin (Neural Networks and Learning Machines) - christopher bishop (Neural Networks for Pattern Recognition)
- axiom 18y agoOy! enough with the neural networks already. Look, a neural network is not some magic machine that can solve all your classification problems. A good 90% of applications of ANNs I've seen could (read: should) have been replaced with a support vector machine, or a Bayesian classifier, or some other proper statistically principled model. I swear, I get the impression that people keep coming back to ANNs just because the goddamn name sounds cool.
- jcbozonier 18y agoSome of us just don't know what they are or how to use them. I guarantee I'll be forming another study group some time in the future on the other topics you mentioned if they are as broad in application as you say. :)
- axiom 18y agoMy advice would just be to take a few stats courses. Make sure you're comfortable with your linear algebra inside out. Learn as many numerical methods as possible etc. etc. In my mind, neural networks should only be used when you're trying to model actual biological systems. Otherwise, one should always use a proper model - since it actually gives insight into what's going on with the system, instead of just being a magic black box. Of course, notice that all commonly used ANN topologies have nothing to do with biology - feed forward, hopfield nets, kohonen maps etc. They have barely any resemblance to how actual neurons behave. Basically, they are just heuristics with a cool sounding name.
- designtofly 18y agoI completely agree with you! The same should be said for genetic algorithms. I don't understand the obsession around here with both ANN and GA. They are interesting tools, but they are not some holy grail with unlimited applications. Both of them are inefficient and rarely the right tool for the job. GA and ANN are used for situations where you don't care about how you got to the solution and what it took to get there, as long as you get close to it in the end.