Where is the line between fitting and actual patterns? - page 29
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
Has anyone looked for a pattern in which direction the wind is blowing? How often have you guessed where it will blow tomorrow? Or in the next hour, hour and a half, 15, 5, 1 minute?
And the wind vane knows... Always...
The wind always has a weather vane at the back.
That's why the weather vane knows something. (insider)
Thank you. I was wondering why you don't use GA on your sample. Well, in MT5 you can make any of your parameters as optimized parameter and use GA. So, we may solve the problem of speed.
GA is used.
But in pre-optimization mode, in case of analysis of other people's MTS, i.e. when you do not understand what logic is laid in it in general.
I'll tell you a secret (in case of robust system) that OOS cannot always be profitable for best sample wholesale
That is, obtaining a loss-making result from OOS we try to go further and reshape filters for supposedly possible loss-model times, overhaul almost completely the TS itself, and so on.
That's right, warmer ))
At the very least, you have to know (test) something to reveal the secret.
Can you, at least veiledly, announce the results of your tests?
joo: применимы ли такие подходы (разделение на Sample, OOS и др.) ко всем ТС без исключения? Если применимы не ко всем, то к каким не применимы?
Professionals of course advise to do without OOS - the more OOS, the faster the TC becomes obsolete. But how to do it in real life - I can't figure out.....
Any TS has the right to life and any of its (TS) rational modification, obtained by adding filters through several years of observation will work in the hands of the creator-trader, if not greedy and give the market a horny elks))
1) results at best optimum (parameter selection is an individual art of the observer-trader)
2) real (profitability within 5-15% of sample time)
3) inevitable future.
Using OOS in real trading = losing profits.
The behaviour of the price after opening a position(in my example Sell) is unpredictable but limited by the probabilities of 4 events. The probabilities are calculated in-sample.
Professionals certainly advise to do without OOS - the more OOS, the faster the TC becomes obsolete. But how to do it in real life - I do not know.....
And could you please tell Reshetov in a private message that he is right.
I'm talking about the extra-gifted....... ;-))
The behaviour of the price after opening a position (in my example Sell) is unpredictable but limited by the probabilities of 4 events. The probabilities are calculated in-sample.
Nice doing business with you.... )
Frankly speaking, I thought that you opened the topic by accident, in a rush....
I confess. I was wrong.
1) results at best optimum ( parameter selection is an individual art of the observer-trader)
2) real (profitability within 5-15% of sample time)
3) inevitable future.
Using OOS in real trading = losing profits.
===========================================================
And please don't be pessimistic.
Plot #3 can be different.....