Machine learning in trading: theory, models, practice and algo-trading - page 1445

 
Искусственный интеллект обыграл сильнейших киберспортсменов мира по Dota 2
Искусственный интеллект обыграл сильнейших киберспортсменов мира по Dota 2
  • habr.com
Технологии искусственного интеллекта вновь заставили о себе говорить. На это раз причиной стала победа OpenAI над командой профессиональных игроков в Dota 2. Название команды — OG, в 2018 году она получила главный приз киберспорта, заняв первое место в турнире The International по Dota 2. Призовой фонд этого турнира составляет $25 млн. На...
 
I have been working as a trader for a long time and I am still working on my account:

A team of some mathematicians at the institute decided to try to exploit their findings in the financial markets. I took it for a lot of money, I won't reveal the details. I am not going to reveal any details.

I have a good relationship with my broker, so please let me know what you think.

 
Ivan Butko:

A team of some mathematicians at the institute decided to try to exploit their findings in the financial markets. I took it for a lot of money, I'm not going to reveal the details. The main thing now is to gain a positive history on the real, but history is history, not an indicator (for everyone, I'm sure).

I do not know if it is the National Market or something else? What do you give for entry?

 
Maxim Dmitrievsky:

Good luck, write about the results sometimes

for sure, if there will be something to show) thank you

 
Aleksey Vyazmikin:

What do you give for entrance?

In the shower for tea, they found a D))) All you have to do is specify the number of layers and neurons. Well, one thing I can say: the more layers and neurons (100 layers, 100 neurons), the more useless (paradox). If I add 2-3 layers of 2-3 neurons, I gain something. Maybe I'm not trained properly - I don't know. Just started, I'll check more.

UPD

Oh, and some kind of "regression" or "classification" to choose from - one of two.
 
Kesha Rutov:

***

Kesha, how old are you?

 


 
Maxim Dmitrievsky:

mbm, I'll finish later and compare what I got without making up anything extra.

In my opinion, the statement about memory loss when going to differences is not particularly important for the further statement. You could just say something along the lines of postulating that memory is only determined by price levels. But perhaps I just don't understand the book business and so I won't insist)

If we proceed from this postulate, then we come to the usual notion of price as a diffusion process in a heterogeneous space of prices. There is, however, an essential difference from physical diffusion - there a particle usually does not affect the properties of the space. The price changes it because when it reaches some new level, the orders set there are triggered and new ones are added at other levels.

 
Aleksey Nikolayev:

In my opinion, the statement about the loss of memory in the transition to differences is not particularly important for the further statement. One could just say something like postulate that memory is determined only by price levels. But perhaps I just don't understand the book business and so I won't insist)

If we proceed from this postulate, then we come to the usual notion of price as a diffusion process in a heterogeneous space of prices. There is, however, an essential difference from physical diffusion - there a particle usually does not affect the properties of the space. The price changes it, because when it reaches some new level, the orders set there are triggered and new ones are added at other levels.

Actually, it says so directly :)

 
Maxim Dmitrievsky:

In fact, it says so directly :)

OK) let's put it aside for clarity)