Machine learning in trading: theory, models, practice and algo-trading - page 2258

 
I don't remember either. Ask the girls around here, they remember everything.
 
Aleksey Mavrin:

Reading the branch (or rather in attempts to read), there is a strong impression that the closed chats with moderation - very useful thing)).

Re the case request - have posted a collection of literature on the subject. If you don't mind the link again please, @Maxim Dmitrievsky,Aleksey Nikolayev

Branch attached to general discussion, 4th from top https://www.mql5.com/ru/forum/214418

Что читать, смотреть и где учиться машинному обучению
Что читать, смотреть и где учиться машинному обучению
  • 2017.08.25
  • www.mql5.com
На русском сайте StackOverflow в вопросе о хороших книгах по математике и машинному обучению сформировали список материалов, с которых стоит начать...
 
Vladimir Perervenko:

The article deals with something else. It considers the case when all predictors are discrete [0, 1]. Then there is a problem. The neural network does not understand predictors with zero variation.

You, as I understand it, have a slightly different case. You have combined in input predictors (continuous) and target (discrete matrix ncol=3). You are trying to get a qualitative distribution of latent from which you generate (restore) the input including the target, practically without training. Did I understand you correctly? It will not work qualitatively. The article shows the way of solution. Translate discrete target to continuous using RBM, connect with other predictors and use BAE (training!). And then retrieve examples from the trained VAE and use RBM to restore the target again. It's a bit complicated. But it might work.

I'll try it with regular AE.

Good luck

you can just train some classifier on this data to give you probabilities.

option even simpler: divide dataset into 2 parts with different labels and train 2 models... and don't bother your grandma with all sorts of conditional states)

 

Tried copulas, coders, tabula gans, codero gans. hmm so far unsurpassed. Copulas are not bad. Neural network technologies are still outsiders for tabular data, which is a shame.

If you need more data, just gmm for now.

 
Aleksey Mavrin:

Reading the branch (or rather in attempts to read), there is a strong impression that the closed chats with moderation - very useful thing)).

Re the case request - have posted a collection of literature on the subject. If you don't mind the link again please, @Maxim Dmitrievsky,Aleksey Nikolayev.

I remembered, that I gave a link to this archive. You can literally read it for a while.

https://codernet.ru/books/python/?page=1

Python | CoderNet
  • codernet.ru
Архив учебной литературы по программированию на языке Python
 

Biomorphic neural network architectures for AI systems

Биоморфные нейросетевые архитектуры для систем ИИ - Вадим Филиппов - Семинар сообщества AGI
Биоморфные нейросетевые архитектуры для систем ИИ - Вадим Филиппов - Семинар сообщества AGI
  • 2020.12.24
  • www.youtube.com
Биоморфные нейросетевые архитектуры для систем искусственного интеллекта следующего поколения: как и зачем? - Вадим ФилипповСеминар русскоязычного сообщества...
 

Hello! Did you run out of Internet?

Happy New Year!

;)
 
Renat Akhtyamov:

Hello! Did you run out of Internet?

Happy holidays!

;)

Everyone was blocked at the same time.

Happy new year to all!

 
Maxim Dmitrievsky:

Tried copulas, coders, tabula gans, codero gans. hmm so far unsurpassed. Copulas are not bad. Neural network technologies are still outsiders for tabular data, which is a shame.

If you need more data, just gmm for now.

Maxim, have you ever tried a Neural Turing machine? In which framework and what were the successes?

Happy New Year and all the best wishes!

 
dr.mr.mom Mishanin:

Maxim, have you tried the Neural Turing machine? In which framework and what were your successes?

Happy New Year and all the best wishes to everyone!

happy new year!!! in siberia it's already come)))))