Elite indicators :) - page 1145

 

I have only seen it very recently and shared it here.The author has couple of more articles about fuzzy models.They can be useful in '' if ,or,then,else'' situations.The code can be handy for market state like '' down trend low volatility , down trend high volatility etc.. I think combining fuzzy indicators with one of machine learning classifiers in an expert may give good resultshttp://www.sfu.ca/~rgencay/jarticles/jbf-fuzzy.pdf

 
nevar:
I have only seen it very recently and shared it here.The author has couple of more articles about fuzzy models.They can be useful in '' if ,or,then,else'' situations.The code can be handy for market state like '' down trend low volatility , down trend high volatility etc.. I think combining fuzzy indicators with one of machine learning classifiers in an expert may give good resultshttp://www.sfu.ca/~rgencay/jarticles/jbf-fuzzy.pdf

Thanks for the link and the pdf

Will be checking it (new ideas are rare these days, all should be checked )

 
wojtek.paul:
Mntiwana, could you please attach the indicators ocn nma from your post #11327? I can't find them. https://www.mql5.com/en/forum/175037/page756

it is here,

ocn_nma_3_3.ex4

Files:
ocn_nma_3_3.ex4  38 kb
 

Thank you!

 
krelian99:
Fuzzy Logic is indeed very awesome, but you need something that works. Have you experience with that indicator from the article? My Russian is so bad.

Krelian,

if only the matter is about language,then you can have in english too.

regards

Files:
1_19.png  228 kb
 

Yes, but sadly the article isn`t in English.

 
krelian99:
Yes, but sadly the article isn`t in English.

dont worry,try this one,:)

https://www.mql5.com/en/articles/1991

Files:
1_20.png  185 kb
 

but dear krelian,it is working strange way,still out of my head,applied different setting.

Files:
3_2.png  116 kb
 

Ok, I read it then in English. Have here so technical probs, so... but Fuzzy Logic it is like humans think: absolutely yes, yes, maybe, no and absolutely no and not only black or white like 'signals'. You divide your areas in these decisions. You have 2 or more, so you must put them together and defuzzy it to yes or no. The taken indicators here are not the best, so with some work and nice setups good results should be possible.

 

What is a problem, is that some arbitrary levels are used for decisions. That is not in the spirit of learning

Reason: