Machine learning in trading: theory, models, practice and algo-trading - page 2656

 
Aleksey Nikolayev #:

We traverse the entire history in the cycle. For each of its points, we look to see if pattern #X has been realised on a prehistory of length N, respectively, this is two more levels of nesting cycles - by N and X).

And this is the very minimum) Mining patterns bruteforce is harsh and merciless, but simple and universal).

But this is not how they program in Rk, they do it correctly without a loop at all.
 
mytarmailS #:
But that's not how you program in Rcpp, it's properly done without a loop.

Well, yes, vector functions, and when you can't find suitable ones (or are just too lazy to understand the package), you use rcpp.

 
Aleksey Nikolayev #:

Yes, vector functions, and when you can't find the right ones (or you're just too lazy to understand the package), you use rcpp.

I envy you so much that you know how to rcpp
 
mytarmailS #:
I envy you for knowing how to rcpp.

It's not easy in our business without c/c++.

 
Hi all. Just found your forum. Can anyone tell me if there is a general advisor here or how this forum works?
 
Aleksey Nikolayev #:

How's it going with the rules? Or is that it?

 
mytarmailS #:

How's it going with the rules? Or is that it?

The problem with the rules (as with any other MO model) is that they can change depending on the interval of history taken for their training. Without at least some well thought out, reasonable algorithm for choosing (and re-selecting moments of) history, everything looks kind of indefinite and a bit ridiculous. Trying to come up with something meaningful on this topic.

 
Aleksey Nikolayev #:

The problem with rules (as with any other MO model) is that they can change depending on the interval of history taken to train them. Without some well thought out, reasonable algorithm for selecting (and re-selecting) history, everything looks vague and a bit ridiculous. Trying to come up with something meaningful on this topic.

this problem is typical for any TA

as several waves can be considered depending on the time interval.

waves overlap and appear depending on the length of trades.

therefore the problem is not solvable

 
Renat Akhtyamov #:

this problem is typical for any TA

since several waves can be considered depending on the time interval

waves overlap and appear depending on the length of trades

so the problem is not solvable

Overlapping wave makers arean extremely disgusting sight.

 
Aleksey Nikolayev #:

The overlapping wave-makers arean utterly disgusting sight.

ahahaha

I agree 100%.

but anyway, it turns out that looking at all the TANalisers is very ..... fun.

Well, there are no fish there, it's high time we realised that.