Mechanisation of optimal parameter selection. Finding a common denominator. - page 6

New comment
 
paukas:

The system is a fit with historical data. If the fit is good, it will be profitable for some time.

If the fit is good, it is not a system.

The second statement is a cliché firmly embedded in the brains of traders who make up the majority in the sad 5/95 statistic.

 
OnGoing:

If it's a fit, it's not a system.

The second statement is a cliché firmly embedded in the brains of the traders who make up the majority in the sad 5/95 statistic.

That's what you think when you're young. In five years it will pass.

That's why smart people invented the forward test. A good fitting system should be able to withstand the forward test in an interval of 1/3 of that at which it was optimised.

 
paukas:
The system is a fit with historical data. If the fit is good, it will be profitable for some time. If it doesn't, it needs to be adjusted again. Trying to come up with a system that works "on all available history" is doomed.
Well yes, it's about assessing how much confidence you can put in a particular fit. The same mat statistics operates with confidence intervals when estimating any parameter and probability is always stipulated. For example, MO with the probability of 0.95 lies in the interval... In trading there are more nuances, but still you have to choose estimates with a sufficiently high degree of confidence
 
paukas:
That's what you think when you're young. In five years you'll be over it.
Not at all, it's the other way round, grey in the beard... I hope that's not your case.)
 
OnGoing:

It is not about what it can be supplemented with, but about the fact that statistics is not an adequate method of evaluation at all. That is why it is called a fitting, a self-deception, an illusion...

If one prefers to indulge oneself in illusions, then one is entitled to do so, of course.

No)))) statistics is a tool and you have to know how to use it properly. To say that any statistic is a hoax is from using it incorrectly)))
 
paukas:

That's why clever people invented the forward test.

So test immediately as smart people do, that's what I'm talking about)
 
Avals:
No)))) statistics is a tool and you need to know how to use it correctly. Saying that any statistics is a hoax is from using it incorrectly)))
Mantra)
 
OnGoing:
So test the way smart people do, that's what I'm talking about.)
That's what I do. And you keep looking for the grail.
 
paukas:

That's what you think when you're young. In five years it will pass.

That's why smart people invented the forward test. A good fitting system should be able to withstand the forward test in an interval of 1/3 of that at which it was optimised.

Forward is a tool which also has its drawbacks. For example, if you use it systematically/multiple times, it loses its meaning and the out-of-sample area becomes in-sample)))
 
Avals:
Forward is a tool that also has its drawbacks. For example, if you use it systematically/multiple times, it loses its meaning and the out-of-sample area becomes in-sample)))
That's true, but the number of trials also grows:)))
12345678910111213...19
New comment