Author's dialogue. Alexander Smirnov. - page 22

New comment
 
ANG3110:
Korey:
to ANG3110.
You have a "poor trader's dream" in your picture, as unattainable as a GAZ-3110
Then you should buy GAZ-T34.

So what is this miracle you have demonstrated?
 

to Mathemat
 Why reconstruct the Smirnov characteristic if it is approximated from the original reference by the Remez method.
It's still a filter. A bit narrower, a bit wider, well, cut it down to twice 6dB/octave..., and you can still look for these octaves.
So about octaves:
The result of this thread - you need real adaptive filters. And what they will adapt
- muving coefficients or Fourier sampling, the result will be the same as we have scrolled on our fingers (in this thread):

Any given linear filter can be matched with a simple sum-difference filter (muwings),
or a superposition of muwings.

 
Yurixx:

So what kind of miracle did you demonstrate?

A simple EMA - three times the run in both directions.
 
ANG3110 писал (а): A simple EMA - three times the run in both directions.
Such perfectly smooth and lag-free curves should be fed to the output of the approximating neural network (as a teacher). It will be something like a smooth analogue of a zigzag.
 
ANG3110:
Yurixx:

So what kind of miracle did you demonstrate ?

It's a simple EMA - three times the run in both directions.

So this is what they call a T3 ?
 
There is a suspicion that the real Alligator is not made for SMMA like all the freebies, but for three backwards.
 
Yurixx:

So this is what they call a T3?


No, that's what I wrote. I don't know what to call it, I call it at_ZDn.

And I wrote about the T3 earlier in this thread. It looks like this:

 
Korey:
There is a suspicion that the real Alligator is not made on SMMA like all the free ones, but on three backwards and forwards.
There is a suspicion that the Alligator "three backwards and forwards" is also not made on real market data, but on specially processed data.
 

It turns out that A.B. Sergienko. I have "DSP", it is the book that the author of this thread suggested to read, but I don't know to whom he recommended it. Mr. Smirnov I recommend you also to look very carefully at your recommended page 216, I enclose the picture. Here it is necessary to make comparison, to input a known signal and to carry out calculations, in this figure a rectangular pulse is applied and at least it is qualitatively visible TF work.

By the way none of the "scientific" terms you introduced are found in this book.


Z.I. Who needs books on DSP, feel free to contact me. Only pliz through skype postage I'll get tired of forwarding them oooooh a lot.

 
ANG3110:


No, that's what I wrote. I don't know what to call it, I refer to it as at_ZDn.

And I wrote about T3 earlier in this thread.

Yes, it's a beautiful and sleek worm! But what's the use in Expert Advisor, because its right end twitches (i.e. changes direction on a flat)?

But even if T3 is equipped with a threshold of some points (the antiflat trigger), it will still make a lot of false reversals because it is very sensitive (in this respect T3 is somewhat similar to Hull МА except for squirts of different amplitude).

Here we have an essential question: does the ТС that outlasts the flat with open positions make sense? After all, if the answer is no, only the slope of the MA makes sense, i.e. there should be 3 colours on the chart.

1...151617181920212223242526272829...44
New comment