expert testing of strategies - page 6
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
Itso, dear man, your thirst for opposition is stronger than logic. Where did I plead the dignity of genetic algorithm and optimizer used in metatrader ? Help answer - didn't try to do that (and certainly not trying to say that your toolkit is a mercer that doesn't work, while mine works), I just said from the start that I haven't used this toolkit, as I've only recently learned about it, and at this point it makes no sense for me to switch to a new one. Maybe they are good, I just don't know it. I was asking a completely different question - do you know HOW they work? Do you know the algorithms (down to the operators) by which the optimization (much less genetic. ...) takes place? You answer this question specifically, I'm interested. Do not accuse me of praising something and criticizing something, I'm not doing it.
Because if you don't know the work algorithm of genetic algorithm and optimizer, it turns out that you are working blindly - it's like using an indicator, which formula you don't know. Selection criteria may be good or bad, but they should be known to the tester accurate to a decimal point and best of all written by the tester himself. I think it is important because in practice, every nuance of the program affects the trading results.
Prival, I think I know which Expert Advisor we are talking about. Let's listen toArtur's opinion. But he needs figures, not a picture. Yes, and the statistics is not enough, generally speaking.
P.S. And thank you very much for Sato. I, too, first prefer the "physics on my fingers" and then the mathematics.
Here's a picture of expert trading on championship quotes, and if this new super method says it's bad, screw its black box. You don't even need the numbers.
Artur - I don't want to be in opposition to you, but your approach is a bit .... confuses me a bit.
You have some criteria and knowledge, (which of course is not bad at all), but you want to sell something (eventually), while not being interested in the existence of analogues of your approach.
In other words - marketing as such does not exist at all.
And I have a surprise for you - we all understand very well the work of built-in optimizer (because we use it regularly), and about the genetic algorithm we know what it gives us - and if necessary we can read 'Genetic algorithms - mathematical tool' and 'Genetic algorithms in MetaTrader 4. A comparison with the optimizer's direct brute force. We know both advantages and disadvantages of the optimizer. In fact, it is the same 'expert estimator' that takes a very long time to make, according to you.
I get the impression that your approach is not mathematical but, shall we say, psychic - (or something like that)?
Dear Itso, I did not intend to sell anything, where did you get such thoughts from? I have written many times here that I am not looking for commercial gain and I have explained the true interest in one of my previous posts directly to the esteemed public.
Not only have you read my posts inattentively, but you speculate about the lack of marketing and reproach me for not being interested in my approach counterparts. That is, exactly half of your post is built on a condemnation of something that no one has suggested.
Regarding the second part of your post, let me draw attention to a couple of details, namely two statements: "I have no idea what it gives us" and "we can read it if we need to" From these phrases it follows that you have not yet read the references that you suggest me to study and consequently you have not studied the tools you use or am I picking on words?
Regarding the question about psychic - I think the question is normal and fair. The answer is that the method is mathematical.
Artur, you please at least give a subtle hint on how you implement this check - "1 - strategy is most likely to fit the curve and has not shown itself to be robust". Thanks in advance!
Artur, let me remind you of my question though.
Integer, the method is a common modification of what is referred to here as linear regression.
Actually my great curiosity was in what points will be earned by EAs of other forum participants, because my own strategies were rated at best with two points, and mainly with one (that was actually the interest of creating this thread).
Surprisingly it turns out that those present have quite advanced results compared to mine. But in parallel I would like to clarify the origin of these excellent results. Whether sufficiently realistic models were used in obtaining them, etc. And in general it seems we have the most pointless thread on the entire forum, solid clarifications of who sells what and who has what marketing. No discussion of the issues.
I have an idea - to initiate another wave of protests and heated arguments. Respected Itso above threw me a couple of links about genetic algorithm. I'll read them carefully (I'll have a look, they seem to be serious works, but I'll look carefully later) and if I find weak points, I'll try to prove on the forum that the algorithm is useless and that it's no more than guessing by coffee grounds; but if I don't find weak points, I'll write that I haven't found them.
I've read the problems described here : '75000 variants - 4GB RAM and 4GB disk cache not enough???' and got horrified... I'm not an easy task to find a robust forex strategy, but instead of searching for it I need to look for errors in other people's programs, and they all work in a strange way... that's why I prefer to write my own programs and not to use somebody else's work. I think the branch on the above link records +1 point in favour of my arguments.