Machine learning in trading: theory, models, practice and algo-trading - page 2355

[Forester]

Maxim Dmitrievsky:

https://www.mql5.com/ru/articles/6351

I don't see a big difference to detrend by EMA, and if we pass some series with different lags, then the idea of fractional differentiation disappears

That is, no better than a delta with a 0 bar. And what sense does Prado see in this complication? I can't believe he hasn't studied the 0-bar delta. Although Gonchar's article doesn't say a word about it. Maybe he really did not investigate it)))

[Maxim Dmitrievsky]

elibrarius:

That is, it is no better than a delta with a 0 bar. And what sense does Prado see in this complication? I can't believe that he didn't study the delta with 0 bar. Although Gonchar's article doesn't say a word about it. Maybe he hasn't really studied it.)

Gonchar just copies his books and showing off.)

Well, the method itself is beautiful, you can choose the coefficient at which the series will remain stationary and the most memorable, sort of

or you can just pick up the MA)

[Forester]

Maxim Dmitrievsky:

Gonchar is just copying his books and showing off.)

He is showing off all right. On the subject he got his education in Italy. He finds foreign clients and earns money on them. It would be more difficult without his educational papers and articles.

[Maxim Dmitrievsky]

elibrarius:
Well, he's showing off all right) On the topic of MO was educated in Italy. He finds foreign clients and earns money from them. Without the education paper and articles it would be more difficult.

He's selling some courses.

For 25 bucks a subscription, that's all the quantum. I am sure there is nothing to read.)

[Maxim Dmitrievsky]

I looked at the amount of material in my articles and the number of original solutions used. It seems that it is already pulling on the book cooler than Pradov's.

[mytarmailS]

If only something else worked, it would be a fairy tale.)

[Uladzimir Izerski]

Maxim Dmitrievsky:
I looked at the amount of material in my articles and the number of original solutions used. It looks like it's already a bigger book than Pradov's.

Go ahead and print the book.

There's nothing to throw around here. On it, by the way, can and earn. But here are taking me into the arms of vague doubts that you will earn, and even more so that write.)

[Aleksey Nikolayev]

Maxim Dmitrievsky:
I looked at the amount of material in my articles and the number of original solutions used. It looks like it's already a bigger book than Pradov's.

Well, he has enough articles too) and patents and a lot of billions under management).

Overall, the book makes a good impression as a statement of a coherent approach to system-building. When it comes to the specifics, however, everything is laid out concisely, cursorily, and seems a bit underwhelmingly fresh and practically useful.

[Maxim Dmitrievsky]

Aleksey Nikolayev:

Well, he has plenty of articles, too) and patents and many billions under management)

Overall, the book makes a good impression as a presentation of a coherent approach to systems engineering. When it comes to the specifics, however, everything is laid out succinctly, cursorily, and seems a little too recent and practically useful.

Something didn't find any foundation under his name. The book is well written and all these suggested tricks are understandable, but the alpha is not there

Most likely it is just written in academic style, professors have to write something

[Aleksey Nikolayev]

Maxim Dmitrievsky:

I couldn't find any foundation under his name. The book is well written and all these proposed tricks are understandable, but the alpha is not there

Most likely just written in the academic style, professors have to write something

I only found a reference that he worked as a chief machinist for half a year at AQR (about 150 billion under management)

Anyway, the book is useful only as a scheme of the process in general.