expert testing of strategies - page 7

New comment
 

Dear Artur, here I looked carefully at your first post, where you say (and I quote)"... At the initial stage it is proposed to use the described assessment "for free". " (end quotation). Such a slip of the tongue suggests a commercial use of your method. You also quite firmly refuse to disclose your method - which of course is your right, but it also suggests commercial use.

I still can't understand your logic - why do you think I don't understand what genetic algorithm is - I gave references for YOU, so you most likely didn't know what is implemented in MT (here "we" includes both me and you).

But all that is trivial. You did give a hint on your method - (I quote) "...method is a common modification of what is called here a linear regression" (end quote). I was wondering what else is called linear regression - can you share some references, literature, etc. Or is it some kind of your development, in which case I am interested to read this and other developments from your pen.

 
Itso, when I started this thread I didn't know what I wanted, that's one. Second, I think that trying to sell a piece of software is almost useless. Large corporations are unable to protect their software products from copying and free distribution through the Internet, and here I am... Yes, after the first sale (if any), the Internet will be flooded with free copies. In general, it's not my business and I'll never do it.
Concerning "why do you think I do not understand what is the genetic algorithm", I'm just trying to convey my position, that without knowing the code of those tools, which you are using, backtesting becomes no more than guessing with a coffee grounds. And it's more like extrosensory magic than mathematics, when results are done virtually from a black box. For instance when I first started using indicators the first thing I did was to find muwiga formula and write it manually - it turned out that it's not just a simple average value that is calculated there, but it's more complicated.
Concerning the linear regression - you can also call it a straight line approximation (or first order polynomial, it doesn't matter) - it sounds more physical, because you are drawing a straight line in the best way through the set of points, but it's a matter of taste how to call it.

And in general I have nothing against what you write, and not against whom here. Our quarrels with you is mostly a filling of pauses, don't you agree? I just find it fun to talk to guys on the forum, see who's thinking and what they're thinking. It's not hard to find a pretext for a heated argument :)
 
Integer:

Prival, what happens to the last sell?


The last one is not a sell but a buy, you can see in my championship profile, I am standing in the buy. This is a variant of the finalised ekpert working on the championship. It's not as smooth as I want it to be.

Edit, there is a small sell before the buy, but it closes as soon as the critical selling point is exceeded

1234567
New comment