| |
| |
| |
|
Page: 1 2
Comments:
<0> hi
<0> are you working on AI stuff, x00q
<0> if youre still here
<1> Well, not /right now/, but in general, yeah
<0> what kinda things?
<1> Lots of things. I've spent (and presently spend) a lot of time working with evolutionary algorithms and neural networks
<1> I little bit with data mining and information compression/retrieval
<1> I used to model corrosive environments with multilayer perceptrons for some material science researchers
<1> ttt-: why?
<0> just wondering :)
<2> DanF_DrC, hey.
<2> So it has been?
<2> DanF_DrC, appear!
<3> i found a cool econometrics page with papers on neural networks: http://www.sinica.edu.tw/as/ssrc/ckuan/index-e.html
<4> <Dark-Elf@ef> wee linear alg >_<
<5> Andares, hehe what can I do for you?
<6> HEY HOW ARE YOU PEOPLE DOING!?!
<5> could be worse :) you?
<6> Good
<6> college is so freaking awesome!
<5> that's good. that only leaves the little problems like war over greed, killing people for oil and money. polution to make money. global warming. we're almost there
<5> but I guess it's good you like college. what about it do you like?
<6> The fact that all the people there are not a total idiot, and I do not have the same cl*** every day x180
<7> i have to be honest
<7> i got along with other human beings on a good level on college
<6> I have finally held a conversion with another human face to face without having to explain every other word...
<7> outside of college people are... f-d up
<7> Instructors who barely speak english were inviting me into their offices
<6> (about computers) I do have other conversaions
<7> outside of college it was just amazing
<6> not often though
<5> hehe I can understand your new found community Shadow_mil
<6> DanF_DrC: well at this point I am still 5 levels above my peers, though its better then before when I was like 50 levels above my peers, we do have some idiots, but most of my computer teachers can understand all I say
<5> :)
<3> i'm glad i haven't gotten in trouble
<6> DanF_DrC: as you know I catch onto things I am intressed in quickly, even if its complex and most people need years and years to understand it
<5> I've certainly seen worse
<8> Shadow_mil, how old are you if you don't mind me asking
<6> 18
<8> ok, thank you
<6> why?
<6> how old are you?
<8> I am 30
<5> me2
<3> also, today i had to sneak into the building because i didn't have my id haha
<6> paros: how old are you?
<6> if you don't mine
<6> _death: where do you go?
<3> anyway, time to get back to the dorms
<6> ....
<8> perhaps he was lagging
<8> Is anyone here actively doing research in an AI field as part of an academic setting?
<6> Me
<6> After my computer programming degree, I wanna double major in AI
<8> Shadow_mil, I hope it works out nicely for you :-)
<8> Anyone else?
<6> eleftherios: I have found it hard to find a place that teachs anything on AI, any suggestions
<6> ?
<8> Shadow_mil, your best bets are curiosity, an open mind, and a pile of books. Some will be good, some will be bad. Take your time and you will fine-tune your reading list eventually. Make sure it covers a wide spectrum, say from cognitive science to cooking.
<6> cooking...
<6> geat I only have to go from cog.... to coo... thats what, maybe 10 books?
<8> Shadow_mil, also, if it is programming you are interested in I think you can't go wrong with SICP and its lectures: http://swiss.csail.mit.edu/cl***es/6.001/abelson-sussman-lectures/
<6> if that.
<6> eleftherios, I don't have to become some type of theroyical scientist do I?
<8> Shadow_mil, who am I to answer such a question?
<8> Shadow_mil, and what is the point of it anyway?
<6> Do I have to go to school for 8 years to learn this stuff
<6> eleftherios: chances are Ill prolly not go to school to learn any of it, might read a few books on it, but Ill prolly comeup with my own modules and break off the main stream AI community, then disapper for 50 years and come back with stuff people have never see or dreamed of.
<8> This I do not know. It is highly dependent on you, the school you will attend, and a myriad other factors which I we are unable to take into account.
<8> Good night.
<6> nite
<2> DanF_DrC, heh, playing highlight tag again.
<2> Are you here _now_?
<5> yep
<2> Okay
<2> Have there been any successful neural networks that aren't just feed-forward?
<6> Kohona* maps
<2> What are those?
<6> Andares SOMs
<6> Self Organizing Mpas
<7> hebbian networks
<7> er
<7> there is another
<7> fully-connected
<7> I tend to think kohonen maps are MORE successful than FF
<2> paros, hebbian networks.. really?
<7> Andares hopfield is another
<7> DanF_DrC do you happen to be familiar with kohonen maps?
<2> paros, what are hopfield?
<2> *networks
<4> <Dark-Elf@ef> paros, any connection to the greek island?
<4> <Dark-Elf@ef> oh on a side note anyone read about that crazy russian mathemagician? :0
<5> Andares, sure. auto***ociation. training unsupervised as a normal neural net and backprop but in the special case; the function input->input mapping and the result is the middle layers
<5> dark-elf, yes
<4> <Dark-Elf@ef> i always find it amusing when people make tremendous breakthroughs then shun everyone
<2> DanF_DrC, theoretically, how would clusters of neurons be organized in the Brain?
<5> theoretically? :) theoretically I don't know. there are some studies of the layout but not a lot on how it is formed afaik. I haven't studied it much but I recall some plates of around 4 layers of neurons. clusters like that then interconnected in larger structures
<5> some odd columns here and there too
<7> something called a 'cortical column' were discovered in mammal brains only
<2> God, it is impossible to find research on the brain through Google.
<7> they start from the outside of the cortex and go straight down in row of about 80 neurons
<7> andares try googling 'spike train encoding'
<2> http://en.wikipedia.org/wiki/Cortical_column
<2> w00t
<2> paros, okay
<4> <Dark-Elf@ef> paros, no connection to the greek island of Paros?
<2> What's a "telencephalon?"
<5> a neuron that has cable tv :)
<2> Ugh.
<2> You may have physically injured me.
<2> With that bad joke
<2> xD
<5> hehe
<5> sorry
<2> The brain seems sort of digital.
<2> But not quite.
<5> might be a dude that handles long distance communication. don't know
<2> http://en.wikipedia.org/wiki/Cortical_column
<2> En este contexto.
<2> Hmm.. I can't believe that the brain could be that complicated.
<2> I mean, it has to have evolved (sorry DanF_DrC), it has to be able to be encoded in DNA, and it has to be constructed :/
<2> Hmm DanF_DrC, "Neural ensembles" seem to be what I should look at.
<7> the brain is also built by experience during life
<2> Of course.
<2> But its structure has to be simple initially, correct?
<7> i suppose its like any animal brain at some stage in the fetus
<2> Do we know how animal brains work any better than we know how human brains work?
<6> yes
<6> No animal a live has as complex decision making, lang, or anyother brain skills as man
<2> But they have some.
<7> Andares we know all sorts of things about mouse brain
<2> paros, really?
<2> paros, how much do we know?
<6> Andares: yes, we can even simulate them
<7> we know the hypothalamus in the mouse is related to memory formation
<2> Mouse v2.0
<7> the amygdala is implicated in fear
<2> Shadow_mil, could you perhaps give me a keyword to search for?
<6> paros: I saw a thing the other day were people were able to use mangets to control a mouse's reactions
<2> Poor mouse.
<7> that's one of the more humane experiments actually
<7> the 'lever experiments' were pretty bad
<2> One thing that doesn't seem well explained.
<2> If a single neuron fires only wen an observer sees his grandmother..
<2> then how can it possibly contain enough information to make that decision?
Return to
#ai
or
Go to some related
logs:
#perl
slackware booting initrd
grogoreo
homespyvideo username
mysql delete identical rows
#python
irc show system stats
#gentoo
#php
perl shorthand variables $*
|
|