Machine learning in trading: theory, models, practice and algo-trading - page 791

[Ivan Negreshniy]

Renat Akhtyamov:
Don't tell me it's real with a total duration of trade more than 3 months

Of course, this is a demo account ... Now we are testing a neural network P-Net, I have already written about it, it is a new development patented in the U.S. and Europe, unfortunately I can not disclose yet, but the roller, not yet commercial (not advertising).

https://www.youtube.com/watch?v=uY4tLXU5Rxc&t=265s

[Vizard_]

Mihail Marchukajtes: I'm waiting for objections to amendments and changes...

I liked it very much! No objections, corrections, changes.
Thanks for the effort, the world shone in new colors! Thank you!

[Mihail Marchukajtes]

Vizard_:

I like it very much! No objections, corrections, changes.
Thank you for your work, the world shone with new colors! Thank you!

I don't recognize you. Are you really you or not you????

[Renat Akhtyamov]

Ivan Negreshniy:

Of course, this is a demo account... Now we are testing a neural network P-Net, I have already written about it, it is a new development patented in the U.S. and Europe, unfortunately I can not disclose yet, here's just a video, not yet commercial (not advertising).

https://www.youtube.com/watch?v=uY4tLXU5Rxc&t=265s

But then what is the point?

[Vizard_]

Mihail Marchukajtes:

I don't recognize you. Are you really you or not you????

I am. It's just if the ideas are genius, I'm not trolling. Tell me something else, please.

[Mihail Marchukajtes]

Here is the content of the article I want to write. the rest tomorrow, so I already have midnight...

Contents

1. Introduction.
2. Analysis and rationale for the choice of direction. Regression or classification.
3. Requirements to the output variable. Basic rules of construction.
4. Subject area analysis and search for the maximum set of explanatory variables.
5. Data preprocessing, search for significant variables for the target function
6. Model training, model list extraction.
7. Basic requirements for determining system performance.
8. Characteristics and preliminary estimation of obtained models.
9. Evaluation of mutual information. Selecting a significant model.
10. Putting the model into operation. Evaluation of the EP site.

[Yuriy Asaulenko]

Renat Akhtyamov:
Don't tell me that this is a real with a total duration of trade of more than 3 months.

There is no difference: real - not real. With adequate modeling and testing, the real is not much different from the test. In fact, the real is absolutely unnecessary for evaluating the system.

The whole question is about the adequacy of the model and the test.

[Mihail Marchukajtes]

Yuriy Asaulenko:

There is no difference: real - not real. With adequate modeling and testing, the real is not much different from the test. In fact, the real is absolutely unnecessary for evaluating the system.

The whole question is the adequacy of the model and test to the real.

The reliable way to ensure the agreement between the real and the test is to work on the first bar without using zero. Checked it myself :-)

[Renat Akhtyamov]

Yuriy Asaulenko:

There is no difference: real - not real. With adequate modeling and testing, the real is not much different from the test. In fact, the real is absolutely unnecessary for evaluating the system.

The whole question is the adequacy of the model and test itself.

See

that's why it's not real

[Vizard_]

Mihail Marchukajtes:

Here is the content of the article I want to write. the rest tomorrow, so I already have midnight...

Contents

1. Introduction.
2. Analysis and rationale for the choice of direction. Regression or classification.
3. Requirements to the output variable. Basic rules of construction.
4. Subject area analysis and search for the maximum set of explanatory variables.
5. Data preprocessing, search for significant variables for the target function
6. Model training, model list extraction.
7. Basic requirements for determining system performance.
8. Characteristics and preliminary estimation of obtained models.
9. Evaluation of mutual information. Selecting a significant model.
10. Putting the model into operation. Evaluation of the CB segment.

Cool. No, write right now, right here, you had an epiphany today...