Hey guys, I'm a computer science student, currently on break from uni, and I was just wondering if you need to be massively good at maths to mess with neural networks? The reason I ask? I'm bored as hell and feel like taking up a challenge.
Just to put things into perspective, I have to do a supplementary examfor computational mathematics, because I fail hard at boolean algebra and karnaugh maps.
Thanks for your time 4chan, I know you'll provide me with the right answer.
Name:
Anonymous2006-12-12 21:55
You don't need to know much math just to implement an NN, but to work with the theory behind it, you will need to to have sharp Vector Analysis and Differential Equations skills. There are some simple but interesting examples of NNs at planetsourcecode.
Name:
Anonymous2006-12-13 9:07
haha oh wow, you suck at math. boolean algebra and karnaugh maps are so easy a high school student could learn them in a week.
Name:
But2006-12-13 9:47
>>3 what' s not easy is to ome up with simple "karnaugh maps" idea. bet my ass no high school student you know could ever do anything like that.
Name:
Anonymous2006-12-13 14:29
>>1
Neural networks contain elements of chaos theory and quantum physics since the electric pulses are continuous, not 1 at a time and each brain cell has it's own properties and can have upwards of 40 connections. You could possibly analyse neural networks and replicate some processes, though 100% certain predictions are impossible.
Name:
Anonymous2006-12-13 18:54
>>4
Come up with "karnaugh maps" idea? I am not talking about inventing karnaugh maps. I am talking about learning it. It wasn't exactly easy to come up with F = ma but plenty of HS students can learn it.
Karnaugh maps are really simple. You do know that they teach calculus in high school. This shit is barely algebra.
Name:
Anonymous2006-12-13 19:43
Welcome to Science & Math where everyone is msarter than everyone else.