Discussion of article "Introduction to the Empirical Mode Decomposition Method" - page 3
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
...The only disappointing thing is that the author has already understood this theory and has an idea how to connect the described methods to practice, but for some reason he kept it silent.
At least a simple implementation in the form of an Expert Advisor or indicator, and then the reader, following the path from practice to understanding the theory, will be able to cope with any material.
When there is a goal (final implementation), the whole path (understanding the theory) is easier.
Imho, you can write a practical article in the form of Part II....
Lumping everything in one pile will make the reader's brain boil... a typical reader...
Imho, you could write a practical article in the form of Part II....
Lumping everything together will make the reader's brain boil... a typical reader...
Somehow it so happened that I periodically had to return to questions related to decomposition. When I got to orthogonal transformations, it was naturally the Fourier transform and the Walsh-Adamar transform. But having realised that non-stationarity of the data under study does not allow to obtain proper results with their help, I had to turn to wavelets. But for some reason they did not bring satisfaction, most likely because I did not pay them due attention. Somehow, once again returning to the topic of decomposition, I came across a mention of HHT.
HHT is an orthogonal transformation that does not impose any restrictions on the sequence under study. It is intended for transforming non-linear, non-stationary sequences. But the most interesting thing is that unlike the previously mentioned transformations, it has no fixed basis, the basis is adaptive. At first glance it is just a dream. But my first attempts to implement the HHT algorithm disappointed me greatly. If I changed even the most insignificant detail in the algorithm, I got completely incomparable results. Therefore, I decided to find a prototype in MQL4 or MQL5 to compare my results with. Having failed to find such a prototype, I abandoned my experiments with HHT.
Some time passed, and another round of experiments led me back to HHT. This time I created an implementation of HHT that already made a slightly better impression. This implementation is what I decided to share by writing an article. The article should briefly touch on HHT itself and propose an implementation of it. The intention was that the article would serve as a prototype for someone else's implementation of HHT, which I was unable to find at the time.
While writing the article, I was very concerned that the HHT implementation proposed in the article might be heavily criticised. I thought I would be flooded with comment questions like:
But the observations turned out to be much simpler and at the same time much more complicated.
I don't know if any further developments are worth sharing?
...
But the comments turned out to be much simpler and at the same time much more complicated.
I don't know if it's worth sharing in the future, any developments?
Don't pay attention to blah-blah from the outside. Anything and anyone can be criticised. Even the most famous members of humanity, who have made billions of dollars, are heavily criticised. Up to the point that their work is called junk. As a rule, such critics are just idiots who can't see beyond their noses, freeloaders who are waiting to be presented to them on a plate or those who are deprived of intelligence by nature and can't do anything worthy of attention.
In my opinion, any experience and research is valuable, even if it is not possible to achieve the desired result. It will "open someone's eyes" and save time, and someone will be pushed on a new path of research in this field. Don't stop - create. ))
In my opinion, any experience and research is valuable, even if it does not achieve the desired result. For some it will "open their eyes" and save time, and for others it will push them on a new path of research in this area. Don't stop - create. ))
Not every experience is valuable, because it is the experience that is valuable, not the game of numbers.
The first sign by which you can distinguish experience from the game of numbers is the very first section, which is usually called "literature review" (from the word steal!). The author, instead of such a substantive section, provides a surrogate for such a section, since Fourier occupies a very specific place in the marketplace - a test input for testing models. At the same time the author ignores the basics in the field of decomposition - Box and Jenkins, and in fact after them a lot of things happened.
That is why I classify this article as a game of numbers, completely disconnected from realities. It is impossible to understand the value of this article, as the place of the considered method among other similar ones is not marked, and also the criteria of result estimation are not specified.
I welcomed the appearance of this article, because on the exceptionally poor statistical field of this forum - this article is a notable phenomenon.
But I express my criticism, because by the standards of a scientific article - it is a game of numbers. But I also express my criticism because I think that the author is not hopeless and will be able to write articles that can be attributed to experience.
My advice to the author of the article is to write, definitely write, and among the critics single out those who contribute to your progress and do not pay attention to those who sit with an open beak waiting for the grail.
...
My advice to the author of the article - write, write, and among the critics, single out those who contribute to your progress and do not pay attention to those who sit with an open beak waiting for the grail.
The critics are on fire! The method is translated into code, is it not enough? Where else will you find everything so chewed up and translated into reality - into code?
Dima, you are the one who is burning up, there was half a step to the indicator, but as a result the article can be used by you and me and you and me :))
I don't need it (I've already told you), I've already written this indicator a long time ago (long before the article). By the way, there are no fish there.
you are the only consumer left :)
Dima, you are the one who is burning up, there was half a step left to the indicator, but as a result the article can be used by you and me and you and me :))
I (I already told you) don't need it, I wrote this indicator a long time ago (long before the article). By the way, there are no fish there.
you are the only consumer left :)
The indicator in the studio - without it all this code is empty text. The topic is interesting.
Critics are on fire! The method is translated into code, is it not enough? Where else will you find everything so chewed up and translated into reality - into code?