Machine learning in trading: theory, models, practice and algo-trading - page 3213

 
mytarmailS #:

how much are you paying?

 
Maxim Dmitrievsky #:

look for yourself

 
mytarmailS #:

look for yourself

I'll give you 10 reasons, and none of them will be statistically significant.

Same thing with your calculations so far.

 
fxsaber #:
Seen several interpretations of the term.

Interpretations from the chattering classes.

Quite specifically.

There is a file A, for example, of size 2000

From it we get file train of size 1500 by generating a random sample by sample()

ind <- sample( 2000, 1500); ind

[1] 830 1537 607 651 1244 211 1033 767 452 744 533 1784 1791 1470 846 1448 354 1304 1274 1901 1991 220 1069 341 730

[26] 591 162 401 583 1296 478 1332 673 491 1264 712 1683 993 509 319 868 798 1398 504 1655 180 1512 1594 1663 575

[51] 745 1263 489 197 1481 1649 387 59 6 34 136 1871 1181 1804 1331 324 1987 1802 1964 783 1260 599 13 1070 1938

[76] 510 1929 1613 944 780 631 414 122 1323 1882 352 1071 838 1473 666 1731 1315 1199 899 686 386 299 483 1351 503

[101] 1588 1475 1738 1346 1636 1359 1670 1054 384 139 719 866 1897 870 361 294 578 772 323 888 964 995 1319 703 1063

.....

train <- A[ind]

test <- A[-ind]

 
mytarmailS #:

So it turns out that a professional trader is a trader who earns money by trading on the market and lives from it.

I know those who meet this definition, you are mistaken with their perception of others.

Forum on trading, automated trading systems and testing trading strategies.

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

mytarmailS, 2023.08.31 14:00

And professionals look at them both with pity and condescension and quietly say to themselves: when will you b...bleep think of changing the concept, not the scenery....

 
Maxim Dmitrievsky #:

I'll give you 10 reasons, and none of them will be statistically significant.

Same with your calculations so far.

Who told you I have one reason and not 1,000?

You made it up, you believed it, you made fun of yourself.... ))

 
fxsaber #:

Those who fit that definition know you are wrong about their perception of those around them.

It's quite possible, I won't argue, it was a joke post, not a statement....

Strange, you claimed 10 min ago that you don't know professionals, and now suddenly you do).

 
СанСаныч Фоменко #:

Quite specifically.

You have failed to explain it clearly. train and test, of course, do not overlap.


Here is a computational process (it is called training, optimisation - it doesn't matter). Are both samples involved in this process?

 
mytarmailS #:

and who told you I had one reason and not, say, 1,000?

I made it up, I believed it, I made fun of myself... ))

Well, you just blurted something out again, without any proof.

 
fxsaber #:

You're not making yourself clear. train and test certainly don't overlap.


Here is a computational process (it is called training, optimisation - it doesn't matter). Are both samples involved in this process?

The code is not clear?

train is a file on which the model is trained, i.e. a list of patterns is formed, if in a random forest, then about 100 patterns.

test is a file where the algorithm predicts the value of the target variable based on the patterns.