Machine learning in trading: theory, models, practice and algo-trading - page 2402

[Alexander Ivanov]

Vladimir Baskakov:
So much has been written, in this time even a vacuum cleaner could be taught to trade. Where is the result?

Hello! It's been created long ago.

https://www.mql5.com/ru/blogs/post/219729

https://www.mql5.com/ru/code/10289

[Mikhail Mishanin]

Evgeni Gavrilovi:

is it realistic to create such an algorithm for the stock market?

https://hightech.plus/2021/04/30/algoritm-mashinnogo-obucheniya-samostoyatelno-izuchaet-zakoni-kvantovih-sistem

Easily.

[Valeriy Yastremskiy]

Evgeni Gavrilovi:

Is it realistic to create such an algorithm for the stock market?

https://hightech.plus/2021/04/30/algoritm-mashinnogo-obucheniya-samostoyatelno-izuchaet-zakoni-kvantovih-sistem

What a load of crap, not an article. Not yet. It's hard to predict for a company with more than 10 bakery workers yet, and for 100 sawmill workers too... And we haven't learned the fate of a child either....

[Evgeni Gavrilovi]

Mikhail Mishanin:

Easy.

can you give me an example?)

[mytarmailS]

Alexander Ivanov:

Hi! It's been established for a long time.

https://www.mql5.com/ru/blogs/post/219729

https://www.mql5.com/ru/code/10289

Ahahahahah...

[Maxim Dmitrievsky]

At the time of '14, he was probably the strongest. At least on this forum :)

[LUIS ALBERTO BIANUCCI]

There are some unresolved issues in the currency or stock market. It is not difficult to write a neural network of the type you need in python or even in mt4, the difference would be the amount of code you need to write. Undoubtedly it would work well initially, but after a while, as market conditions change, it will need to be optimized, more necessary to optimize the network every time there is news against the trend, stop working after it and optimize it the next time. The risk of over-optimization. It follows that neural networks should stop a few hours before important news, and then they should self-optimize, after the news, a few hours after the news, but here the programming is extremely difficult.

[Viktor Zhuravlev]

LUIS ALBERTO BIANUCCI:

There are some unresolved issues in the currency or stock market. It is not difficult to write a neural network of the type you need in python or even in mt4, the difference would be the amount of code you need to write. Undoubtedly it would work well initially, but after a while, as market conditions change, it will need to be optimized, more necessary to optimize the network every time there is news against the trend, stop working after it and optimize it the next time. The risk of over-optimization. It follows that neural networks should stop a few hours before important news, and then they should self-optimize after the news, a few hours after that, but here the programming is extremely difficult.

Thank you! You're the one we've been waiting for 2400 pages!!!

[pips]

Victor:

Thank you! You're the one we've been waiting for 2400 pages!!!

We had a laugh, yes. No faith whatsoever))

[Mikhail Mishanin]

Evgeni Gavrilovi:

Can I give you an example?)

An example of what?