Hmmm, you can do any linear separable function with the artificial neuron, not just AND logic function. And hidden layers network are supposed to model any non-linear function. The frequency-rate firing approach is also interesting, if you didn’t investigate this way :).
My projects are under development, I still need some components to make my ideas become real, but some tests are positive. I started writing the pages, but I need more content and some photos … coming maybe next week !
mate im not a digital freak hahah ofc theres thousands ways to rach any target. and i tried to make an intelligent system without any processor myself too.(which had instinct,memory,random patterns,ect for a simple 4 legs creature which could learn.) but that time i had no idea how a neural network works (well i dont know much now either) so i end up adding a microcontroller and then refused to make it yup its extremely interesting and wannable to put so much effort to ur own ideas.
aaaahh theres lots to talk about this subject.u know this reminded me of MARK TILDEN hahaha.
one last thing im not a true programmer but i like to try every method.an i believe that the feature belongs to organic systems.we’ll miss pure electronics
yes please do share ur project here with us.ill try to find my papers too.LMR folks FTW!
nah, that AND logic table just shows how it functions.
oh and “the frequency-rate firing approach”, do you meen the weight of every neuron? i used capacitors for those.
and btw, this robot won’t really be something like a very smart robot, heck no! i don’t have funds for that (my parents won’t allow me to buy parts), this robot will only be a way to show me if my theories about my neural network would work, so that when i have some parts, i can make a realistic one.