Discussion of article "An Introduction to Fuzzy Logic" - page 2

 
Vladimir Perervenko:

What about the rules? Let's see the article.

Good luck.

Well the rules for linking input and output terms are self-explanatory.
 
Vladimir Perervenko:

PS. By the way, I've been meaning to ask you for a long time. Would you like to transfer your product written in Java to R language?

The question is not in the right place. CRAN is responsible for porting to R, and I have nothing to do with this organisation. Moreover, I don't know R and I'm not going to spend time on learning it, because I'm not interested in visualisation of games in numeric.

At the moment the question is whether it is worth spending time on mastering fuzzy logic or whether we are dealing with another duck. If you read boastful articles about fuzzy logic, it turns out that almost every self-respecting financier uses it. If we look at the tools used by financiers for decision-making, it turns out that the most common is Excel with various packages, among which fuzzy logic packages are clearly not observed.

After all, it is one thing to formulate a set of rules for a stationary area where everything is immutable and therefore formalisable and workable even with tolerances and landings plus/minus a kilometre. It is another matter to spend time on formulating rules in the non-stationary domain, where everything changes with every new tick and eventually everything will have to be rewritten from scratch. The rules have to be written and debugged manually. You have written them, debugged them, and then some aunt from some central bank changes the interest rate, and all your work is for naught.

 
Yury Reshetov:

Wrong question. CRAN is in charge of R porting, and I have nothing to do with this organisation. Moreover, I don't know R and I'm not going to spend time on learning it, as I'm not interested in visualisation of games in numerics.

At the moment the question is whether it is worth spending time on mastering fuzzy logic or whether we are dealing with another duck.

The modern level of applied mathematics is such that there is a huge amount of mathematical methods in the form of ready-made code, documented, with examples..., literature.....

So what?

That's not the problem!

The problem is not to slip into a game of numbers, not to be like that monkey who tried on glasses.

The basic rule of statistics says: rubbish in, rubbish out.

Like in a meat grinder: what you put in, you get out, but in a different form.

Applied to the article.

After all, the main question is: for what areas in trading is this fuzzy logic applicable?

The ARMA model was invented 40 years ago. At the same time they clearly specified a number of nuances with stationary and non-stationary series. It turned out that a step was made in the financial markets, but a very small one with huge dangers on the sides. And what did the author tell us about the applicability of fuzzy logic? Maybe I did not catch something?

The caret shell of R includes over 150 packages from the field of machine learning. So what? We know a lot about the applicability of each of the packages of this shell on the financial markets, but to earn something you have to work a lot with input data, at least 70% of the time is spent on input data, 10% on the application of the ready-made, debugged algorithm. The rest is spent on trading techniques, including risk management. By the way, there is no fuzzy logic among the caret shell packages.

Hence, the main efforts in the field of fuzzy logic (as well as any other mathematical tools) should be focused on the formulation of the conditions of APPLICABILITY of this tool in the financial markets and in particular at Forex. And at leisure, if you wish and have time, to understand the internal structure of matmetod.

PS.

Just for reference from R

fuzzyFDRExact calculation of fuzzy decision rules for multiple testing
FuzzyLPFuzzy Linear Programming
FuzzyNumbersTools to Deal with Fuzzy Numbers
fuzzyRankTestsFuzzy Rank Tests and Confidence Intervals
FuzzyStatProbFuzzy Stationary Probabilities from a Sequence of Observations of an Unknown Markov Chain
FuzzyToolkitUoNType 1 Fuzzy Logic Toolkit
CRAN - Package FuzzyLP
  • cran.r-project.org
Carlos A. Rabelo
 
СанСаныч Фоменко:

Hence, the main efforts in the field of fuzzy logic (as well as any other mathematical tools) should be focused on the formulation of the conditions of APPLICABILITY of this tool in the financial markets, and in particular at Forex. And at leisure, if you wish and have time, to understand the internal structure of the matmetod.

However, you are spreading your lips.

First deal with the internal structure. And then at your leisure, if you have free time, you can meditate on its applicability to trading.

Jokes, jokes. But that's the way it is in any business. I.e. a lot of time will be spent on making bumps, stepping on rakes. And only when it becomes known where the most painful rake lies, something will start to work out.

If we look at fuzzy logic as a kind of automatic regulator like PIDs, which only smoothes out the jitters compared to its competitor, then who knows how it can be applied in trading? It is not enough money to regulate the market. On your own depot only the balance is controlled, and equity dances wherever it wants. But I don't need balance - it's for schoolboys. Maybe regulate risks? But here the percentage of the depo rules a lot and without any jerkiness, not to mention that you don't need to make up any rules, because it is set by only one constant.

I don't know, where could this odd one be screwed in?

 
WeChat free experience shouting, another profit more than a thousand points of QQ group: 375124107, plus group please note "77", thank you for cooperation!
 

As clearly stated in the title, this is an introduction to fuzzy logic, but that's very rough introduction, don't expect to fully understand it if you don't already know what is fuzzy logic.

The second part is commented code, with as usual examples not related at all to trading.

 

Hello, I dropt script on chart, and give me this error:


 
Milad Nadi:

Hi

when i updated metatrader to build 2342

all of samples with fuzzy logic library 

return error "incorrect casting of pointers" on MQL5\Include\Math\Fuzzy\RuleParser.mqh Line 712  

please help to fix bug 

many thanks

Forum on trading, automated trading systems and testing trading strategies

New MetaTrader 5 Platform Build 2340: Managing account settings in the Tester and expanded integration with Python

Sergey Golubev, 2020.03.02 12:00

Hi @Milad Nadi:

If it is related to those codes (Fuzzy - library for developing fuzzy models) so you shoulkd ask the author/coder to update this library (MetaQuotes has nothing to do with it sorry. The website of the author/coder are published in the beginning of this discription: Fuzzy - library for developing fuzzy models

 
Milad Nadi:

hi

when i updated metatrader to build 2342

all of samples with fuzzy logic library 

return error "incorrect casting of pointers" on MQL5\Include\Math\Fuzzy\RuleParser.mqh Line 712  

please help to fix bug 

many thanks

read my previous post.
 
Sergey Golubev:
read my previous post.

hi 

thanks for your replay

real author is Dmitry Kalyuzhny. but his coded this library in .Net

But maybe someone else changed the code to mql

i see the author of this article is Дмитрий Калюжный

but I could not find Дмитрий Калюжный profile on mql5

Would you please help me to contact or fix this bug 

many thanks