|
where do i start
|
#
?
Aug 21, 2017 00:44
|
|
|
|
| # ? Apr 25, 2024 16:46 |
|
|
i've been loving around with keras for the last month or two. makes it pretty easy to start getting results without caring too much about the math underneath. you'll need to learn the math eventually though.
|
|
#
?
Aug 21, 2017 01:48
|
|
|
sounds like some fotm bullshit to me op
|
|
#
?
Aug 21, 2017 01:50
|
|
|
Disappointing Dollhouse posted:i've been loving around with keras for the last month or two. makes it pretty easy to start getting results without caring too much about the math underneath. I wanna learn the math
|
|
#
?
Aug 21, 2017 01:51
|
|
|
it depends what you mean by these things op. when it was released the game fear was praised for having great AI, but the "AI" was just the A* search algorithm with heuristics. there's some good stuff in the bots thread which might give you some ideas for idiot projects for machine learning w/ neural networks. really the math isn't that bad, but it depends what your goals are or what you're interested in specifically, the words on their own are now meaningless buzzwords
|
|
#
?
Aug 21, 2017 01:57
|
|
|
go to the tensorflow site and do the tutorial there https://www.tensorflow.org/get_started/get_started#tensorflow_core_tutorial
|
|
#
?
Aug 21, 2017 02:01
|
|
|
it's vector calculus and linear algebra all the way down the rest is figuring out how to approximate approximations to statistically mostly correct estimators of models that are vaguely inspired by hilariously simple neuroscience garbage
|
|
#
?
Aug 21, 2017 02:15
|
|
|
good poo pooThe Management posted:go to the tensorflow site and do the tutorial there nice
|
|
#
?
Aug 21, 2017 02:36
|
|
|
learn to make better posts first, op
|
|
#
?
Aug 21, 2017 03:15
|
|
|
Skim Milk posted:learn to make better posts first, op i'd markov chain your post history but i doubt it'd be an improvement
|
|
#
?
Aug 21, 2017 03:22
|
|
|
qhat posted:where do i start https://www.coursera.org/learn/machine-learning here u go
|
|
#
?
Aug 21, 2017 03:24
|
|
|
qhat posted:i'd markov chain your post history but i doubt it'd be an improvement dont schizo-shame
|
|
#
?
Aug 21, 2017 03:41
|
|
|
gotta start with regular ol human learning
|
|
#
?
Aug 21, 2017 03:43
|
|
|
yo this is fully sick im watching all of this poo poo right now
|
|
#
?
Aug 21, 2017 04:11
|
|
|
neural networks are matrices with racism inside
|
|
#
?
Aug 21, 2017 04:19
|
|
|
Bloody posted:neural networks are matrices with racism inside looking forward to running a clustering algorithm on crime statistics and seeing what it has to say about demographics
|
|
#
?
Aug 21, 2017 04:28
|
|
|
last night i learned about naive bayes classifiers and how to make a linear regression model while touching up on my linear algebra that i haven't done in over a decade. good website.
|
|
#
?
Aug 21, 2017 20:38
|
|
|
qhat posted:naive boys
|
|
#
?
Aug 21, 2017 20:58
|
|
|
echinopsis posted:keep that poo poo outta the pos 
|
|
#
?
Aug 21, 2017 21:07
|
|
|
this is actually pretty cool op if only i wasnt so lazy though
|
|
#
?
Aug 21, 2017 21:49
|
|
|
[quote="“Powerful Two-Hander”" post="“475599117”"] this is actually pretty cool op if only i wasnt so lazy though [/quote] same that’s why i pay money to internet course men and tell all my friends so i can shame myself into doing it
|
|
#
?
Aug 21, 2017 21:58
|
|
|
Bulgogi Hoagie posted:same that’s why i pay money to internet course men and tell all my friends so i can shame myself into doing it it's free
|
|
#
?
Aug 21, 2017 22:00
|
|
|
[quote="“qhat”" post="“475599494”"] it’s free [/quote] i only really did edx courses for stuff and you can pay there if you want to get a shiny linkedin cert, it’s not very expensive
|
|
#
?
Aug 21, 2017 22:05
|
|
|
Bulgogi Hoagie posted:i only really did edx courses for stuff and you can pay there if you want to get a shiny linkedin cert, it’s not very expensive LinkedIn certs are worthless and nobody cares about them.
|
|
#
?
Aug 21, 2017 22:06
|
|
|
A LinkedIn cert is like the gold sticker the teacher gives the div kid in school because he was able to correctly recite his 2 times tables. Nobody cares about it.
|
|
#
?
Aug 21, 2017 22:10
|
|
|
qhat posted:LinkedIn certs are worthless and nobody cares about them. my autism cares
|
|
#
?
Aug 21, 2017 22:11
|
|
|
just have the machine learn about machine learning for you, op. work smarter, not harder etc
|
|
#
?
Aug 22, 2017 00:41
|
|
|
Read this book, op.
|
|
#
?
Aug 22, 2017 01:21
|
|
|
lol no the book you want is called "deep learning"
|
|
#
?
Aug 22, 2017 01:34
|
|
|
i'll put the deep in deep learning
|
|
#
?
Aug 22, 2017 02:30
|
|
|
Deep Dish Fuckfest posted:lol no the book you want is called "deep learning" I haven't read that one. The one I linked I had done in undergrad. It was a pretty straightforward book in my opinion. Is it too simple? Is it off wrt real world stuff? I'm genuinely interested.
|
|
#
?
Aug 22, 2017 02:48
|
|
|
yo I want to make a computer vision system that I can teach what <object I need to count> looks like then show it a picture and have it count the number of the objects in it. I've heard about YOLO for things like this but I'm not sure it applies. what should I use?
|
|
#
?
Aug 22, 2017 03:24
|
|
|
op: no you don't if you waited this long for it to be "cool" then you are just chasing a trend and you're not up for the math maybe what you actually want is to be able to bullshit sufficiently about these topics to put them on your resume?
|
|
#
?
Aug 22, 2017 03:26
|
|
|
solid advice from the jewkiller
|
|
#
?
Aug 22, 2017 03:28
|
|
|
i did a little genetic thing to find trade routes in Elite Dangerous and it was a good learning thing also its mostly resume fodder but if someone ever asks about genetic algorithms ill be ready!!
|
|
#
?
Aug 22, 2017 04:47
|
|
|
JewKiller 3000 posted:op: no you don't Nope, I want to do it properly.
|
|
#
?
Aug 22, 2017 05:13
|
|
|
why? what is your motivation for learning this thing? what if i told you that behind all the fancy math, it is mostly horseshit, and people will move on to the next thing within 5 years or so once all the promises don't pan out?
|
|
#
?
Aug 22, 2017 05:21
|
|
|
qhat posted:where do i start kraftwerk
|
|
#
?
Aug 22, 2017 05:27
|
|
|
qhat posted:where do i start with Lisp, read Norvig then read Lenat's papers "The Nature of Heuristics," "Theory Formation by Heuristic Search," and "Eurisko: A Program That Learns New Heuristics and Domain Concepts" as well as the papers about RLL-1 which goes a bit into the implementation details, like how the frames were set up eschaton fucked around with this message at 06:21 on Aug 22, 2017 |
|
#
?
Aug 22, 2017 06:12
|
|
|
|
| # ? Apr 25, 2024 16:46 |
|
|
you could also try Understanding Artificial Intelligence it has some stuff about the TI Explorer in it!
|
|
#
?
Aug 22, 2017 06:26
|
|