Machine learning in trading: theory, models, practice and algo-trading - page 1190

 
mytarmailS:

interesting articlehttp://www.long-short.pro/post/indikator-haosa-i-rezhimy-fondovogo-rynka-886

And the blog in general.

This is a trend indicator. It depends on the ratio of drift to variance, but can be uncorrelated with them individually

 
Aleksey Vyazmikin:

Have you tried expressing his indicator in code?

no

 
mytarmailS:

nope

Why not?

 
Aleksey Vyazmikin:

Why?

It's not quite clear how))

 

Who can tell me how such systems are created?


 
mytarmailS:

Who can tell me how such systems are created?


What is there?

a standard terminal, the standard panel "Trade in one face" and some kind of indicator with arrows, the window mosaic clicked

all this comes as a standard MT, what is so unusual?

 
Igor Makanu:

What is there?

The usual terminal, the standard panel "Trade in one face" and some kind of indicator with arrows, the window mosaic clicked

all this comes as standard with MT, what is so unusual?

)))))

The question is how to create robots like on the video, to keep up with the market and take the whole trend without any left signals

 

the RL manual from the"reinforcelearn" package

https://markusdumke.github.io/reinforcelearn/index.html

Man, my head is exploding from the number of variables, so many things and all not understandable((

Reinforcement Learning • reinforcelearn
  • markusdumke.github.io
With you can set up a reinforcement learning agent to solve the environment, i.e. to find the best action in each time step. The first step is to set up the policy, which defines which action to choose. For example we could use a uniform random policy. In this scenario the agent chooses all actions with equal probability...
 
mytarmailS:

)))))

The question is how to create robots like the ones in the video, so that they keep up with the market and take the trend without any wrong signals

Ahh, I see, I was too lazy to watch the video, I rewound it in three clicks without sound ))))

Well, apparently it's an advertisement for an advanced bot )))), as an option, use Maxim's libraryhttps://www.mql5.com/ru/code/22915 and after optimization it will work like clockwork... but only on history, I think you do not need to explain that even the future does not know ;)

RL algorithms
RL algorithms
  • www.mql5.com
Данная библиотека имеет расширенный функционал, позволяющий создавать неограниченное количество "Агентов". Использование библиотеки: Пример заполнения входных значений нормированными ценами закрытия: Обучение происходит в тестере...
 
Igor Makanu:

aaaah, I see, I did not watch the video - lazy, in three clicks rewound without sound ))))

Well, apparently it's an advertisement for an advanced bot)))

This is not an advert, but a live broadcast signals in online mode

Reason: