Machine learning in trading: theory, models, practice and algo-trading - page 1056
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
Variant 1: without transformation of traits. Learning from 08.01 Anything before OOS
expected
expected
let microlouche mclay throw the gmdh portrait - it won't be expected. Now on the randome transforms through cosines I will look at
The levels were taken only 4 pieces to start with:
Don't distract the magicians. Sorcerers will show)))
Black (High-Low)-actual, blue-tren, red-forecast...
Maestro, you're back in time - or I already had a desire to run away from here.
But I'll work some more. I feel sorry for those who suffer, myself included. I'll just take a little time out, read some stuff and drink some alcohol.
Here is with the transformations of the signs. Slight decrease in the error on the OOS, as a consequence on the OOS for 10 days in the +, then rolled down again
But my transformer is, to put it mildly, so-so.
In RL cv, or the old one?
the old, oob.
did cv - it doesn't learn anything at all then ) and longDon't distract the magicians. Sorcerers will show)))
Black(High-Low)-actual, blue-trend, red-forecast...
Oh my god, volatility prediction gurus, bow toAlexander_K and others....
What could be more complicated))
Alexander I have to delete my picture to look cool and mysterious? .... it's really a circus
Banderovite?)))
it starts....)))
One thing I can say is that Koldun does not work with the first differences on the ticks.
He cooks, like a cook, his own BP.
Moreover, in his opinion, the expectation of this BP at any time for any sample should be = const, most likely, according to Kolmogorov = 0.
Yes, this is a requirement for stationary VR and he obviously does it somehow.
In the standard case, without tricks, we have the wave packet price "walking" like this:
And so, if someone makes such a BP so that, at least, the expectation is = const (God forbid with variance), he can prepare dustbags. Ugh, you - for dust. I said it all.
One thing I can say is that Koldun does not work with the first differences on the ticks.
He cooks, like a cook, his own BP.
Moreover, in his opinion, the expectation of this BP at any time for any sample must be = const, most likely, according to Kolmogorov = 0.
Yes, this is a requirement for a stationary VR and he obviously does it somehow.
So what's the result? Where's the result?
So what is the result? Where is the result?
Aleshenka, for example, says that he cannot show the state under the contract with investors. For some reason I believe it.