Discussion of article "Grokking market "memory" through differentiation and entropy analysis" - page 3
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
The article is great in terms of the style of presentation and the material in it. It was very pleasant to feel like a complete zero even after trying to understand the algorithms of the source...
However, there are always connoisseurs of "goodness" who have deep knowledge to evaluate the efforts of "younger comrades".
It's a pity that paths with such super-specialists cannot but cross. To the non-existent black list, in short.
Hee)))))) Nice one
Help, error compile
Auto_optimizer.mqh
'virtual_optimizer' - function already defined and has different type Auto_optimizer.mqh 47 18
47 "CAuto_optimizer::virtual_optimizer(void) {"
fractional_entropy_trader.mq5
can't open "C/////D0E8209F77C8CF37AD8BF550E51FF075\MQL5\Include\Auto optimizer.mqh" include file fractional_entropy_trader.mq5 13 11
i use
https://www.mql5.com/en/code/1146 - aglib
https://www.mql5.com/en/code/16006 - mt4 orders for mt5
Help, error compile
Somebody decided to try to run it after all.
There are two applications in the article.
One is theoretical. There, indeed, anything can be substituted.
The other is practical: Tester. You can't do that. You just destroyed the expectation matrix with this method.
But it is through the Tester that you analyse prospects. And it shows such a picture that you will end up throwing a lot of good ideas in the bin.
ZЫ We need some kind of competent instruction on how to use the Tester.
is this a request to me? or do you mean the standard tester :)
Why can't you do it? You can run it on all ticks, the picture won't change much. It's just that the cloze is involved in the training, yes, but nothing prevents it...
By the way, you can make it even easier. Take any cast. symbol with a periodic function, for example, as here and see how the whole cart will go. It will go approximately the same way as on the graph. So the optimiser is OK, the problem is in the data.Is that a request to me? Or do you mean the standard tester :)
Why can't you do it? You can run it on all ticks, the picture won't change much. It's just that the cloze is involved in the training, yes, but nothing prevents it.
Yes, about the standard one. I'd turn the spread to zero for tests like this. When the spread is defeated, only then you can start increasing it.
Besides, there are limit orders, which would significantly improve mat. expectation (now it is zero with a thousand trades).
Yeah, the standard one. I would turn the spread to zero for such tests. When the spread will be defeated, only then you can start increasing it.
Besides, there are limit orders, which would significantly improve mat. expectation (now it is zero with a thousand trades).
Ah, I see. Above I added how the algorithm was tested in general, i.e. whether it picks up patterns. The rest is a matter of data, yes.
I have other TS a little bit, this one is like an unploughed field, open for suggestions on what can be added to it
Somebody decided to try and run it after all.
I ran it, I decided not to comment on syntax errors in the code, it is difficult to assume why the underscore sign was missing in several file names, it would be good to reupload the source code if possible - I think this issue will recur from time to time.
on the article. the author definitely has a talent to explain complex things in an accessible language, once again I am impressed by it - THANK YOU!
Although I like to read Maxim's articles for the fact that they are compact and in essence (without water), but in my opinion this article should be divided into 2 parts, one part is the implementation of the entropy indicator and some tests or statistical data (I did not like the choice of TF M15, for some reason I got the impression that on this TF was just a better result? ..... you need to check in general) and the second part of the article on the auto optimiser, you could try to choose the time of trading or some "market context" to analyse
the author of the article has given very great and interesting material for research.
THANK YOU again!
Valaki még mindig úgy döntött, hogy megpróbálja futtatni.
hi
thx help
solved 50%
--------------------
'#property' - semicolon expected fractional_entropy_trader.mq5 6 1
'MT4_ORDER' - declaration without type MT4Orders.mqh 181 16
'MT4_ORDER' - declaration without type MT4Orders.mqh 215 16
'const' modifier not allowed for nonmember functions MT4Orders.mqh 248 27
'}' - expressions are not allowed on a global scope MT4Orders.mqh 283 1
Thank you in advance
I ran it, I decided not to comment on the errors, it is difficult to assume why the underscore was missing in several file names, if possible, I would like to reload the sources - I think this question will be periodically repeated.
on the article. definitely the author has a talent to explain complex things in an accessible language, once again impressed by it - THANK YOU!
Although I like to read Maxim's articles for the fact that they are compact and in essence (without water), but in my opinion this article should be divided into 2 parts, one part is the implementation of the entropy indicator and some tests or statistical data (I did not like the choice of TF M15, for some reason I got the impression that on this TF was just a better result? ..... I would like to check it in general) and the second part would be an article on auto optimiser, you could try to choose trading time or some "market context" to analyse
the author of the article gave a very great and interesting material for research
THANK YOU again!
I thought it would not be interesting to cite all sorts of statistical studies, screenshots of which I showed in other topics on the forum. Decided to put everything in one pile without fine details. Also, I didn't have time to pick up TC - hyperparameters can beadded (for example, what Alexander wrote that you need to pick up the window - all this is put in hyperparameters), but it's long so I didn't write it all here. Well, I wrote above in the posts how to improve for closing prices - improvements will really be, for example, the threshold-filter on entropy improves even on clozes. Eurobucks m15 is just often used. It showed good results on the DAX. I have not looked at the others. I pointed out the problem that now a linear trend is presented as a feature, it will not work in all situations.
It was not my goal to post a ready robust algo, but to show a possible direction.
I don't know about errors, I've uploaded my working sources.