Machine learning in trading: theory, models, practice and algo-trading - page 2601

[Maxim Dmitrievsky]

Valeriy Yastremskiy #:

To each his own))) Or rather to taste and color))))

How do you test eurofunt from 2009 in 4? Or it is easier to drag on the 5?

What prevents you from the lack of quotes? If you change your brokerage company and increase the maximum number of bars in the window (in the settings), maybe you should try it.

[Valeriy Yastremskiy]

Maxim Dmitrievsky #:
What prevents, not enough quotes? If you change your brokerage company and increase the maximum number of bars in the window (in the settings), maybe you should try it.

I have a question where there are the quotes from the 9th year. I have only found one from February 21, the quotas are also not much. The number to the maximum, no quotes.

[Maxim Dmitrievsky]

Valeriy Yastremskiy #:

Yes, the question is just where are the quotes from year 9, I only found from February 21, quotas also not much. The number to the maximum, no quotes.

as for mt5, they have them on the metaquot server since 1993, mt4 has not been used for a long time

[Valeriy Yastremskiy]

Maxim Dmitrievsky #:

MT5 has been on the metaquot server since 1993, mt4 hasn't been used for a long time

I should have downloaded the quote archive. It's from the 7th year).

[Maxim Dmitrievsky]

How to install CatBoost on the m1 arm

it works, checked it out.

native installer for python is promised soon

[mytarmailS]

How to distinguish randomness from regularity...?

For example, if you search through a huge data set, there is a high probability of finding a combination of parameters that is formed by chance, but looks like a pattern...

How do you distinguish a random non-regularity from a real pattern?

[Valeriy Yastremskiy]

mytarmailS #:

How to distinguish randomness from regularity...?

For example, if you search through a huge data set, there is a high probability of finding a combination of parameters that is formed by chance, but looks like a pattern...

how do you distinguish a random non-regularity from a real pattern?

Only by identifying a cause and effect relationship. If there is no cause, it's sb.

[Aleksey Nikolayev]

mytarmailS #:

How to distinguish randomness from regularity...?

For example, if you search through a huge data set, there is a high probability of finding a combination of parameters that is formed by chance, but looks like a pattern...

How do you distinguish a random non-regularity from a real pattern?

The question is not sufficiently formalized (mathematically). What exactly is meant by the absence/presence of a pattern/randomness?

There are no rigidly fixed and unambiguously defined notionsof regularity/randomness in mathematics.

[mytarmailS]

Valeriy Yastremskiy #:
Only by identifying the cause and effect relationship. If there is no cause, then it's sb.

It is not always possible to trace causation

[mytarmailS]

Aleksey Nikolayev #:

The question is insufficiently formalized (mathematically). What exactly is meant by the absence/presence of regularity/randomness?

In mathematics, there are no rigidly fixed and unambiguously defined notionsof regularity/randomness.

Specific example:

There is a logical rule that predicts something with 80% probability on a trayn and on a test and crossvalidation passed, etc.. And on validation data (absolutely unknown new data) the rule works at random level...

There is another rule which behaves exactly the same way as the first one on a tray, test and passes validation also without problems, those are real regularities...

Question: How can I distinguish one rule from another at the stage of train, test, crosvalidation... Before the validation stage...

I wonder if there are some signs to distinguish between one and another, maybe some statistical tests for randomness or determinism, etc...