Adaptive digital filters - page 10
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
I would love to help. But unfortunately I can't read MQL code as freely as MathCad where formulas are written the way we are used to seeing them in books. The only thing that seems to me (though I'm not sure) is using one of regression types, to make it clearer
There is a linear regression like y(x)=ax+b. You can calculate coefficients a and b in different ways, you can use ANC (seems not to be used there), and you can use recursion, but to understand it you should clearly understand cycles (I get confused there, where, what and why is calculated). Most probably there is a non-linear regression, because there are some if() while calculating + type of regression equation itself is not clear, how many coefficients there are.
In general, almost all indicators can be considered as digital filters, the MA is a digital filter. The word adaptation usually means that some parameters (coefficients in the filter gut) have to change depending on the characteristics of the input signal. Therefore first of all I would refer AMA, FRAMA and similar adaptive digital filters (averaging parameter (n) changes depending on input process variance estimation), almost all FFT, wavelet filters that use threshold processing (trying to match TF parameters with a spectrum of input desired signal).
But SATL, FATL are not adaptive, because TF coefficients were calculated once at design stage to match the transient response of the filter with the spectrum of the input signal (AFR and IFR), and during operation these coefficients do not change. These are the so-called matched filters. But there is an ideal, what is called in DSP optimum filter, to build it is difficult, but possible. For this you need to know spectra of useful signal and noise.
I don't know, if I helped you or confused you :-), but anyway good luck.
And me and Mathemat and someone else saw this noise on ticks. Moreover, on the ticks it is clear that +-1 points have higher probability of the reverse movement than its continuation. Unfortunately, this regularity is inside the spread. And it is not high.
And the fact that it appeared after the processing is interesting.
So you have a deep detrending and should be left with a return, since you can see the pips.
When they talk about measurement noise they mean random deviation of measured data from true value of measured quantity. For example, radar (for specialists :-)) gave range value 105 and true value 100, in next measurement 99 instead of 101, etc. The distribution of the error is generally normal. In case the price comes in, e.g. 1.2567 - this is its true value, the error is zero! What kind of noise are we talking about?
Why can't we actually use the term "the true price value"? You may agree that it is as inaccessible for us as a true range value for radar :). Then there is indeed a difference: "radar" has to "hit" the true range value, while it is enough for us "to hit" the measured price value. But we can make the hypothesis that the true price value is more suitable for prediction than the measured one, and this hypothesis is as good as all the others, explicit or implicit, underlying any other MTS.
I'm a bit confused. If returns is a discrete series of +-1 pips, then mine is more accurate, the output of the filter gives an estimate, i.e. the result in fractions of a pip.
When they talk about measurement noise they mean a random deviation of measured data from the true value of the measured quantity, e.g. a radar (for experts :-)) gave a range value of 105, while the true value was 100, in the next measurement 99 instead of 101 etc. The distribution of the error is generally normal. In case the price comes in, e.g. 1.2567 - this is its true value, the error is zero! What kind of noise are we talking about?
Why can't we actually operate on the concept of "true price value"? It is just as inaccessible to us as the true range value for radar :). Then there really is a difference: "radar" needs to "hit" the true range value, while we only need to "hit" the measured price value. But we can make a hypothesis that the true price value is more suitable for prediction than the measured one, and this hypothesis is as good as all the others, explicit or implicit, underlying any other MTS.
It's not a hypothesis, it's a fact. You have to predict the "true value", if you predict the measurement, then I've laid out the result in pictures here 'Tick collectors. Optimization. DDE in VB (VBA)'.
And before making any predictions, you should try to measure as accurately as possible, because prediction errors are directly related to measurement errors. That's why the further away the forecast is from the measurement point, the worse (less accurate) it is.
Also, "1.2567 is its true value, the error is zero." There is no such thing. It is a measurement of that "true" value that no one knows. It is simply the same as saying that this particular brokerage house knows the true price. And all other participants of Forex, who do not use this brokerage company's data, can think the other way. Suppose that deutsch bank at this time will think that this price is 1.2566. Who is right, where is the truth?
Prival, the truth is where you work. If your brokerage will have a quote of 1.2567 and no quote of 1.2566 (from Deutsche), then no confidence intervals will help you.
Your reality is strictly limited to your brokerage company. You will be allowed to open only at 1.2567 - even if this quote is beyond your favorite confidence interval, meticulously built using data from 100 different brokerage companies. And you cannot raise any objections to your brokerage company on the basis that it is an outfitter, because it defines the rules.
1.2567 is an exact measurement that is absolutely real for you (because you can open at this price).
Prival, the truth is where you work. If your brokerage will have a quote of 1.2567 and no quote of 1.2566 (from Deutsche), then no confidence intervals will help you.
Your reality is strictly limited to your brokerage company. You will be allowed to open only at 1.2567 - even if this quote is beyond your favorite confidence interval, meticulously built using data from 100 different brokerage companies. And you cannot raise any objections to your brokerage company on the basis that it is an outfitter, because it defines the rules.
1.2567 is an exact measurement that is absolutely real for you (because you can open at this price).
I don't know, maybe I'm not explaining it right. The fact that I can only open at 1.2567 at this moment in time, yes absolutely agree. But it's crazy to use only this number to predict IHMO and assume it's true and accurate.
But to use only this number for prediction IHMO + to assume it is true and accurate is nonsense.
I couldn't figure out what was bothering me, but it turns out this sentence. The probability of events occurring may not be equal to one. I see, it all depends on the point of view.