Probabilistic neural networks, packages and algorithms for MT4 - 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
YuraZ, so you are in the championship and you do not give up yet ? Give me a link to your fighter! About the adaptation and retraining once there was a discussion. I suggested to write a simple system. For example by stochastic or crossing two wipers. And then just try to optimise it every day for a not very long period. I don't know if anyone has carried out such an experiment. It is difficult in the tester. And it's even harder on the demo. This was just before I disappeared from here and forex in general for almost a year. So, where to look for it now - I don't know. But the difference between the grids and simply optimized expansions is purely quantitative. Grids simply have more parameters, but they do not contain any ideas. By the way, I still do not understand how to deal with completely ineffective grids. On the one hand it is not good because ideas save computational resources and input information. On the other hand it is good because a strong idea still needs to be found. So, the question is not so simple. But optimization and reoptimization are common that unites both approaches.
https://www.mql5.com/ru/users/YuraZ/
:-) I'm not draining yet... I'm 25th in the table of ranks crawled up to 10th place - it was a record of this expert (among Russian developers 5th place now)
of course he can not win, but he should finish positively in this I believe
I set the bar to increase the deposit up to 40k - for him it will be a record!
there is no adaptivity and elements of NS in the Expert Advisor. he got 80k in training in 2007 01 01 2007 - 20 08 2007
later on, after improvement of the indicator which did not pass the 5 minute boundary! the system raised 120k over the same period without 3 orders limit to 140k
but it was after September 20th and i couldn't send it anymore
i think it is very difficult to write a system with an adaptation - i.e. nc - it is very complicated!
i have already developed an adaptive indicator element - i hope to use an adaptive Expert Advisor next year
I hope to participate next year - an adaptive Expert Advisor - if I have time I will write a neural network
as for the standard indicators - I only use the moving averages - by the way, BETTER also uses moving averages
I am a supporter of my own developments ...
write a leaky system? why?
what about the NS - they have no ideas . - i think you are wrong ... the NS has ideas BETTER proved ...
Look https://www.mql5.com/ru/users/Better/ it's a probabilistic NS
Statistica. It's new. If you need it, I'll send you the link. Free ;) How much more.
Statistica. It's new. If you need it, I'll send you a link. Free ;) How much more.
I would be very happy to. Thanks in advance :) mine is логин@mail.ru
Statistica. It's new. I will send you the link if you need it. Free ;) What more to learn.
Also:
para-b@ya.ru
Here's a link to a forum where a nice person called klot has been learning neural networks and sharing his experiences in the process and the source code. Thanks to him for that.
Since I have some general idea of neural networks and came to them not by fashion, but to implement my own ideas after thinking over various other options, it was enough to read the whole thread for one day and create and run a simple neural network expert the next day, not having any experience with NN at all until that moment.
here is the link: http: //www.fxexpert.ru/forum/index.php?showtopic=656
The toolkit given there makes it easy to create and train a neural network (using Neuroshell2). This network later can be either easily used via MT4 external library or generate a small (actually very small) C source code that is easily ported to the MT4 source code.
There are references to literature in this thread.
I wish everybody have a good time while studying neural networks and please don't be lazy to read the whole article, in order not to overload the branch with questions that have already been answered fifteen times.
Me too. My inbox is in my profile. Thank you.
Statistica. New, I need a link to murman-trans@rambler.ru thanks!
Statistica. It's new. If you need it, I'll send you a link. Free ;) How much more.
3. Showing off :)
A file on gold