Discussion of article "Genetic Algorithms - It's Easy!" - page 6

[Andrey Dik]

hrenfx:

I realise that it is very important to assess the validity of the fit. One method is to add noise to the original data.

The source codes of alternative optimisation methods can be found here(http://alglib.sources.ru/optimization/) and here(http://ool.sourceforge.net/).

Obviously, each optimisation algorithm performs better on its own classes of target functions.

Which target functions do you use in practice?

When developing and debugging UGAs, I used the test target functions in the branch you pointed out. They are different in the nature of their landscapes - there are both smooth and with sharp breaks. UGA handles all of them.

[hrenfx]

joo:
When developing and debugging UGAs, I used the test target functions in that branch you pointed out. They are different in the nature of their landscapes - there are both smooth and with sharp breaks. UGA handles all of them.

I phrased my question poorly. What optimisation problems are you solving, as applied to financial markets? I realise that you can solve anything. I am interested in what you solve.

[Andrey Dik]

hrenfx:
I did not formulate my question well. What optimisation problems do you solve in relation to financial markets? I realise that anything can be solved. I am interested in what you are solving.

I mostly use an algorithm for training neural networks as part of my ATS.

[hrenfx]

Prival:

It's just that your second task (its) solution, is necessary for solving this task here

https://www.mql5.com/ru/forum/123072/page6#254964 (very interesting thread by the way)

At one time I wanted to (combine these two problems) and calculate, look, think, but the hands as always did not reach (time as always in the day is not enough).

Look here.

[Andrey Dik]

Updated UGA library and examples to the article. Current free author's version 1.3.0.

[Andrey Dik]

Update Library UGA and examples for this article. Current free authoring version 1.3.0.

[Andrey Dik]

Updated UGA library and examples to the article. Current free author's version 1.3.1.

I express my gratitude to shurick, mrProF, Serj_Che, Urain for found errors, valuable advice and wishes. You have helped me a lot to make the algorithm better, thank you.

[Andrey Dik]

Update Library UGA and examples for this article. Current free authoring version 1.3.1.

[zsazsa]

Great article! Thank you. 

As I can see you wanted to use UGA for optimizing the training of NNs. Was it successful? What were the genes of the chromosome in this case? Have you developed your own NN library as well or have you found any existing NN implementation supporting the integration with MT5 and GA?

[Serge]

Gentlemen! If it's not too much trouble, please give me some examples of using the UGA library.