Artificial Intelligence General - Peter Norvig Edition
>LINK OF THE DAY
http://fastml.com/numerai-like-kaggle-but-with-a-clean-dataset-top-ten-in-the-money-and-recurring-payouts/
>Fun Video
https://www.youtube.com/watch?v=qv6UVOQ0F44
Now with Discord server.
https://discord.gg/010Hh6pWkFkbMgPyG
Come and discuss, learn about, and panic over all things artificial and (marginally) intelligent.
Links:
>Deep learning frameworks
https://www.tensorflow.org/ (google)
http://caffe.berkeleyvision.org/ (Berkeley Vision and Learning Center)
http://deeplearning.net/software/theano/ (Université de Montréal)
http://www.cntk.ai/ (Microsoft)
http://torch.ch/
>Frameworks comparison
https://en.wikipedia.org/wiki/Comparison_of_deep_learning_software
>Required reading
http://www.deeplearningbook.org/
>Intro videos for beginners:
https://www.youtube.com/channel/UC9OeZkIwhzfv-_Cb7fCikLQ/videos
>Required watching
https://www.youtube.com/watch?v=bxe2T-V8XRs&list=PL77aoaxdgEVDrHoFOMKTjDdsa0p9iVtsR
>Something else that you should probably watch
https://www.youtube.com/watch?v=S_6SU2djoAU&list=PLy4Cxv8UM-bXrPT9-ay4E1MuDj1KFTg9H
https://www.youtube.com/watch?v=6oe1Tmg9rjM
>Books
magnet:?xt=urn:btih:a680ca553dd69a6a5e8b7dc8c684ceb006c7ecc5&dn=Artificial%20Intelligence
>Other Resources
https://facedetection.com/
http://opencv.org/
http://kaldi-asr.org/
http://sirius.clarity-lab.org/
Has anyone ever worked on a pattern recognition problem where you have 90k classes? I'm thinking of using k-Means clustering to reduce it to 500 but I'm not sure on how to proceed on building 90k/500 classifiers to classify them into the corresponding clustered class...
>>54731468
Finally something worth reading on this fucking board
>>54731468
>somebody else started the thread for once
Now I can get comfy.
But just so you know the discord link is only valid for like 24 hrs, so it'll expire at like 10am tomorrow.
Hot
http://arxiv.org/abs/1605.06640
http://arxiv.org/abs/1605.07157
bumo
>>54731483
Have you tried the typical dimensionality reduction techniques, ie PCA and ICA?
>>54731483
>>54736818
Sorry, misread your problem wrong. If you have that many labels then first I'd ask what the distribution is like. If one or two make up the majority of clarifications, then I'd keep those ones then merge all the others into some "Other" label.
>>54731468
>http://www.deeplearningbook.org/
That books formatting is all fucked up senpai
How is the case studies in lisp book? I downloaded a pdf a while ago and never got around to reading it.
>>54736844
That's what I did, I merged all the classes for which I had less than a 100 samples, but I still got 70k labels or so.
>mfw when I'm running kMeans on apache spark with 6 cores
>mfw it has been running for 33h and it's now barely finishing
>>54736922
It is a nice, practical introduction into the field of AI. It doesn't go too deep into theory so I would read his other book for that instead.
>>54731468
>all links are machine learning
>picture only covers classical AI
>>54736922
http://www.norvig.com/Lisp-retro.html
tl;dr good lisp book but outdated as an ai text [and this conclusion was made 20 years ago]
>>54738530
>>54738554
Alright, then have something more modern then:
https://a.cocaine.ninja/wbvopk.pdf