Adaptive digital filters - page 38

New comment
 
mikfor:
Mislaid , if the man had a grail, he wouldn't be on this forum. The 99% are losers here. And the 1% with mixed success. Like me.

Got a little confused with the subject matter of the forum - 99% here are people who learn programming. The other 1% are those who help them.

And don't just jump to the conclusion that you are above 99% of everyone else. It's ridiculous.

 
You need not just an inertia-free swing, but one that turns only in case of an upcoming new movement in which the price passes at least a certain number of pips, allowing for a non-pips profit. But I think these are only unrealizable dreams:))) Who would undertake to design such a machine:)))
 
Yeah, and to send the money to the account herself.
 
khorosh:
You don't just need an inertia-free wrench, but one that turns only in the event of an upcoming new movement in which price passes at least a certain number of pips, allowing for a non-pip-sized profit. But I think these are only unrealizable dreams:))) Who would undertake to design such a machine:)))
you need two wands... and not short... but long.... on 15mTF about 1500-2500 period... somewhere over there.... and then after the crossing, the price passes at least a certain number of pips, allowing you to make a non-pips profit
 

mashka, like the limit of dreams :)))

.

And if the car is adaptive... and a digital one...

.

there is a category of people who are fascinated by beautiful but unintelligible terms :))))

.

.

but watch the news here

 
mikfor:
Sorento , put out for the public's pleasure such an indicator, which does not redraw, does not lag, and even REVERSE by 0.5 bars.

And the keys, too, you boorish man, on a platter?

DDD

Clever, for you, and for the amusement of the public - a quote.

The Levinson-Durbin iteration procedure has shown very good convergence, which proves the fact that fluctuation of EUR/USD is a time series of autoregressive type and is generated by the following recursive relation

where {f(n)} is the normalized white noise and the normalizing factor b 0 is chosen so that the first component of the vector equals 1. This is a very important side-effect of the spectral analysis from which it follows that a one step forward prediction filter is possible for the EUR/USD exchange rate fluctuation time series.
;)
 
Sorento:

Clever, for you, and for the amusement of the public, a quote.

Using "good convergence" to determine whether a process is autoregressive is complete nonsense.

 
lea:

Using 'good convergence' to determine whether a process is autoregressive is complete nonsense.

No, in principle the conclusion is correct - the quoting process is indeed autoregressive with some certainty. But there is no way to predict anything based on it, at least by known methods, due to very low RMS. Because of this the variance of the estimate of the future value of the series is so large that it reduces the benefit of prediction to almost zero, while the remainder is eaten up by the spread. This has already been verified in a lot of papers, I'm too lazy to look for it...
 
khorosh:
You don't just need an inertia-free swing, but one that turns only in the event of an impending new movement in which price passes at least a certain number of pips, allowing for non-pips profits. But I think these are only unrealizable dreams:))) Who would undertake to design such a machine:)))

Why don't you consider the long-period mash-ups?

here, we already have one... now add another one... with a shorter period...... and this EA will give more than 70% of correct inputs.....


 
Aleksander:

Why don't you consider the long-period mash-ups?

here, we already have one... now add another one... with a shorter period ...... and this advisor will give more than 70% of correct inputs.....



there are no outputs according to your methodology, and badly picked trawl is suicide for ATC
1...3132333435363738394041
New comment
Reason: