Machine learning in trading: theory, models, practice and algo-trading - page 771

 
Maxim Dmitrievsky:

Are there no source codes in the pluses? So you can copy and paste them, and then figure them out as you go along

In C++? It's possible that everything is written in Python, the library statsmodel. So far I haven't been able to figure it out. I can't grasp the point of how the array is created when debugging it.

I may abandon all this for now.

 

Who can suggest what about the classification of the price range, what is the best type of network or something else?

And also about R and Python. For example, how to test a neural network? I feed the quotes and get the weights of the network. Or do you use a script to make a connection?

If you do it with the script, then the optimization will probably take a very long time.

 
forexman77:

In C++? It's possible that it's all written in Python with the statsmodel library. I haven't been able to figure it out yet. I can't grasp the point of how the array is created when debugging it.

Or maybe I'll give it all up for now.

Yeah, but I haven't found it myself.

but I did come acrosshttp://www.forexmt4.com/mt_yahoo/ARIMA.mq4.

Try

 
Maxim Dmitrievsky:

Yes, but I couldn't find it myself.

But I did come acrosshttp://www.forexmt4.com/mt_yahoo/ARIMA.mq4

try

There is some kind of flea as a result)

 
forexman77:

There's some kind of a flea-bitten guy in there as a result)

Yes, it does not draw well... old code, apparently.

 
forexman77:

Only after reviewing all the theories and clips, came to the conclusion that the same thing, as a verse memorized say, like parrots.

Exactly. Hundreds of articles on the Internet about Arima, and everywhere - "find autocorrelation and autoregression", then a dozen pictures, and immediately the answer with three parameters without any explanation. In 10% of articles may even mention that there is seasonality.

I took a picture from a neighboring topic, vitally -

 
Vizard_:

Dear Teacher, you are a wonderful Trickster. I understand that Maksimka, who's stoned, can't understand
the meanings of the posts, often losing the thread... but you! Misha, I didn't suggest anything, I only asked
Your Majesty to show me a couple of pictures, that's all!))) hilarious...

I did not do it myself, above post guys posted a picture on my data according to your method. Kind of like...

 
Dr. Trader:

Case in point. Hundreds of articles on the Internet about Arima, and everywhere - "find autocorrelation and autoregression," then a dozen pictures, and immediately the answer with three parameters without any explanation. In 10% of articles may even mention that there is seasonality.

I took a picture from a neighboring topic, vitally -

Well, used to approach things seriously) So I want to understand how it works. Theories are fed up, little understanding of them anyway, it's all in the "bird's tongue". They probably even do it on purpose, so that no one understands anything. So they won't ask how to do it in practice, because they themselves do not know.

 
Today the fate of the villain has again slipped a mishap. Today I switched from daylight saving time to daylight saving time, I had to change the time zone, and I knew this trick, but I understood it only in the evening. Fixed in time, but got a small minus on the turn. It's good that I realized it in time... Everything seems to be working now... The robot is set up and moving on. Today it already had a couple of profitable trades, so the minus was not so bad. The main thing is that the current positions were closed with profit and everything will be fine. Moving on....
 

Well, in many ways you are right. Theory and practice are as different as 0 is different from 100. That is, it is diametrically opposed. Once you learn the theory and move on to practice, you begin to wonder just how much of a difference it is.

I am a practitioner. My goal is to make money in the market, not to prove that one approach or another is right. Or to prove to myself that this method is the best for me. I will take any method, which brings profit. Because my task is to EARN money, and not to spin forever networks, forests, mpeli, etc. convincing and disappointed in these or those methods.

Set yourself a goal to start earning and start working solely from it....

The difference between the robot and the latest version is huge. And tfu,tfu,tfu now he seems to do all the trades that the AI requires of him. So my advice to you, go from theoretical research to practical trading with minimal deposit and minimal risk but on real money and you'll understand how much different theoretical research from practical trading....

P.S. Achievements in theory will not bring you money. Only practical trading and only it...

Well, that's just it... Just thinking out loud... inspired by....

Reason: