Machine learning in trading: theory, models, practice and algo-trading - page 215

 
Yuriy Asaulenko:

214 pages is a lot to study/learn. (And everyone is talking about something different and not always understandable).

Is it possible to summarize all these pages in one post, even if not very short? Type: set goal, methods of solution, results, conclusions.

Let me say right away that my model of the market is a random process (Brownian motion), or rather the sum of several (maybe many) such motions with feedbacks. And it is absolutely useless to predict or look for any regularities except statistical ones. That is, any meaningful predictors simply do not exist, at least for speculative purposes.

It is impossible to summarize as there are a lot of participants in the branch, everyone has his own understanding, his own vector of research, his own truth...

about randomness and Brownian movements of all kinds...

Think objectively...

The exchange/market is a business created by very influential and rich people and as any business it is designed to make money, every market movement is money.

It is very naive to think that in the whole ecosystem, where every movement is money, which was created for the sole purpose of making money, there is even a little bit of chaos, very naive ...

The share of chaos is the share of our misunderstanding of the process I am sure in it...

 
ivanivan_11:
Of course, put the code. it's interesting

R code in mql I can't

Do you know R? Will this code be useful to you?

I'd rather give you the link https://www.mql5.com/ru/forum/86386/page192с where I told you about my idea and try to use mql for your own code

but if you still need it after I said it, I'll give you the code in R

the idea is in fact elementary and understandable
Машинное обучение: теория и практика (торговля и не только)
Машинное обучение: теория и практика (торговля и не только)
  • www.mql5.com
Добрый день всем, Знаю, что есть на форуме энтузиасты machine learning и статистики...
 
mytarmailS:

R code in mql I can't

Do you know R? Will this code be useful to you?

I'd rather give you the link https://www.mql5.com/ru/forum/86386/page192с where I told you about my idea and try to use mql for your own code

but if you still need it after the above mentioned, i'll give you the code in R

the idea is elementary and understandable
Just post the project in R)) Thanks
 
mytarmailS:

It is impossible to summarize, because there are many participants in the branch, each of them has his own understanding, his own vector of research, his own truth...

about randomness and Brownian movements...

Think objectively...

The exchange/market is a business created by very influential and rich people and as any business it is designed to make money, every market movement is money .

It is very naive to think that in the whole ecosystem, where every movement is money, which was created for the sole purpose of making money, there is even a little bit of chaos, very naive ...

A fraction of the chaos is a fraction of our lack of understanding of the process, I'm sure of it...

Big money equals big movements and long deadlines (time intervals). Big money is made on such movements and timelines. The other movements are ripples in the water. And this ripples, this is our interval - the interval of medium and small speculators.

A fraction of the chaos is a fraction of our lack of understanding of the process I'm sure of it...

As for chaos, even if it is not chaos, for us it will always remain chaos. Similar to guessing the next song on a radio station. It may already be predetermined (or it may not yet be), but it's only in the DJ's head. And for everyone else, the choice is random.

 
ivanivan_11:
just a project in R)) thanks

first file, train the models, and save them, also save a new date with clusters from each model

second file, create a target and look for a good combination of clusters

then the new data are recognized by the saved models and we look for that good combination in them (the new data)

I wrote all the code for me, so just ask if you don't understand something

Files:
1.txt  3 kb
2.txt  1 kb
 
Yuriy Asaulenko:

Big money equals: big movements and long timelines (intervals of time). Big money is made on such movements and timelines. The other movements are ripples in the water. And this ripples, this is our interval - the interval of medium and small speculators.

A fraction of the chaos is a fraction of our lack of understanding of the process I'm sure of it...

As for chaos, even if it is not chaos, for us it will always remain chaos. Similar to guessing the next song on a radio station. It may already be predetermined (or it may not yet be), but it's only in the DJ's head. And for everyone else the choice is random.

For me personally there are no movements, there are only levels from which the price bounces, but this is my truth, my imho, I will not impose anything
 
Yuriy Asaulenko:

214 pages is a lot to study/learn. (And everyone is talking about something different and not always understandable).

Is it possible to summarize all these pages in one post, even if not very short? Type: set goal, methods of solution, results, conclusions.

Let me say right away that my model of the market is a random process (Brownian motion), or rather the sum of several (maybe many) such motions with feedbacks. And it is absolutely useless to forecast or look for any regularities except statistical ones. That is, any meaningful predictors simply do not exist, at least for speculative purposes.

Then what are you doing here? Or are you alluding to a transcendent intuition? About spirituality that cannot be formalized...
 
mytarmailS:
personally for me there is no movement, there are only levels from which the price bounces, but this is my truth, my imho, I will not impose anything on anyone

Either it bounces or it punches).

Exactly to the contrary.) There are no levels from which price bounces, there is only movement. The movement is life (just kidding). We work on the movement, and we should be looking for the movement, and not a stop (level). Likewise, not imposing, or even proving. It is an ideology, like yours.

 
Renat Fatkhullin:
  • added a powerful function ArrayPrint for automatic printing of arrays, including structures
  • added FileLoad and FileSave functions for fast writing/reading arrays to disk

Metacquotes is like watching me. And they do everything I need them to do (while stealing grails instead (just kidding)). But with a delay.

A long time ago started saving snapshots of glasses in binary files. Two hundred gigs accumulated.

I don't want to change anything, everything is saved in string and clearly readable.

FileLoad .

The function allows you to quickly read data of a known type into an appropriate array.

Confused by the irrelevant word "fast". Fast as a bullet).

Isn't it faster than reading a binary file in series through FileReadDouble() and putting the read data into an array...?

 
I'm on finmachine today. Maybe I'll learn something new. The program is rich, mostly about banks.
Files: