Machine learning in trading: theory, models, practice and algo-trading - page 2505

 
Rorschach #:

Random walk from random walk random walk.

Random walk from random walk random walk.


There is a direct relationship/conversion between the ACF and the spectrum. Taking the stray spectrum is not obvious, for 1/(f^2), using the increments, the spectrum flattens out, from the side phase is rotated by 90gr.

No, it's much simpler and the spectrum has nothing to do with it at all. You just need to honestly write out the definitions of SB and ACF and calculate by them.

About spectra (there are two types, which are often confused) will be the next question)

 
Aleksey Nikolayev #:

There is a problem that I sometimes propose to local mathematicians to solve. The answer is usually a set of rudeness and swearing. Can you break this sad tradition and answer the essence of the question?

Problem: to calculate the ACF of a random walk.

You can read that to understand how all this is not clear)))

http://hsehelp.ru/sites/default/files/БИ/3%20курс/Эконометрика/лекция_18_эконометрика.pdf

ZS Russian letters in the url address is evil))))) got only so. Highlight and jump to the page. Insert a link or a copy of the link is not translated correctly).

 
Valeriy Yastremskiy #:

You can read it to understand how it all doesn't make sense)))

http://hsehelp.ru/sites/default/files/%D0%91%D0%98/3%20%D0%BA%D1%83%D1%80%D1%81/%D0%AD%D0%BA%D0%BE%D0%BD%D0%BE%D0%BC%D0%B5%D1%82%D1%80%D0%B8%D0%BA%D0%B0/%D0%BB%D0%B5%D0%BA%D1%86%D0%B8%D1%8F_18_%D1%8D%D0%BA%D0%BE%D0%BD%D0%BE%D0%BC%D0%B5%D1%82%D1%80%D0%B8%D0%BA%D0%B0.pdf

The link does not define the ACF of the SB, but there are expressions for the ACF of white noise and the ACF of the stationary AR(1).

There's also a definition of sampling ACF there (fits any series, including SB), but sampling ACF and ACF are completely different things.

Keep searching)

 
Aleksey Nikolayev #:

No, everything is much simpler, and the spectrum has nothing to do with it. You just need to honestly write out the definitions of SB and ACF and calculate by them.

About spectra (there are two types, which are often confused) will be the next question.)

Are you following in Prival's footsteps and want to analyze vibrational links?

Maybe replace the acf with a Hearst?

Or is this purely a sporting interest?

UPD

https://studfile.net/preview/10706509/#3


You can still take it out of the spectrum. For white noise, the ACF is a delta function, and the spectrum is uniform at infinity (coupling in your face, spectrum from delta function). By analogy, you can get the ACF from the SB, the inverse of the PF from 1/(f^2). But this is for lazy people who don't want to count.

 
Rorschach #:

Are you following in Prival's footsteps and want to analyze oscillatory links?

How about replacing the acf with a Hurst?

Or is this purely a sporting interest?

No, there is no applied value here - he has been carrying around this pointless task for a long time.

And strongly wonders why no one pays attention to it and isn't going to dig through textbooks to find the answer.

 
Rorschach #:

Are you following in Prival's footsteps and want to analyze oscillatory links?

How about replacing the acf with a Hurst?

Or is this purely a sporting interest?

Basically, I ask this rather elementary question on the forum only to those who generously and unreasonably throw around on the forum mate. terms) I do not really understand why it makes some forum members so nervous)

As soon as I see a theoretical sample distribution of Hearst for SB, allowing to estimate the deviation of prices from SB so I will immediately replace. All serious studies I've seen of Hearst for real prices do not give it a confidence interval excluding 0.5

Just keeping up the secular near-mathematical conversation)

 
Aleksey Nikolayev #:

In principle, I ask this rather elementary question on the forum only to those who generously and unreasonably throw around on the forum mate terms) I do not really understand why it makes some forum users so nervous).

As soon as I see a theoretical sample distribution of Hearst for SB, allowing to estimate the deviation of prices from SB so I will immediately replace. All serious studies I've seen of Hearst for real prices do not give it a confidence interval excluding 0.5

Just keeping up the small talk about math)

The most accurate results I've come across based on wavelets

 
Rorschach #:

UPD

https://studfile.net/preview/10706509/#3

It does not say anything about the ACF of the SB (only about the selective ACF) And it is not quite correct about the ratio of ergodicity and stationarity.

Rorschach #:

You can still deduce it from the spectrum. For white noise, the ACF is a delta function, and the spectrum is uniform at infinity (coupling in your face, spectrum from a delta function). By analogy, you can get the ACF from the SB, the inverse of the PF from 1/(f^2). But this is for the lazy who don't want to count.

There is a spectrum of realization of a random process - it is a random function (for SB it is defined). There is also an energy spectrum, which is obtained by Fourier transform from ACF for stationary process (generalized to quasi-stationary). SB has no energy spectrum because it is neither stationary nor quasi-stationary. When they say SB spectrum (they also say Brownian or brown noise), they mean a "tweaked" version of SB, which is a stationary process.

 
Aleksey Nikolayev #:

Mostly just to keep the conversation going.)

Proficiency (in the sense of mathematics) is too loud for all of us here) maximum - some meaningfulness of statements)

I.e. to show off again and bring destructiveness to the topic and its logic and take up someone else's time... matrix modeling is for deterministic processes, but not stochastic ones... random wandering is our main asset (that's where you end up, apparently), and derivatives have their distributions... which is why you are wandering around the thread, dreaming of saying something clever... ...and you look smart and doodle your crossword puzzles... preventing people from approaching the topic by burying them in your off-topic libel...

(after all, the question of other people's lives is somehow very pointedly raised in this thread... on someone else's hump?)

p.s.

And the mispricing of derivative assets can already declare an investor's risk-aversion... this is where the price of the underlying goes to look for a new balance - NOT randomly, but logically! goes...... there is only a question of timing... well, in addition to the eternal question of cheap/expensive and in what conditions...

 
Aleksey Nikolayev #:

There is no about the ACF of SB (only about the sample ACF) And also it is not quite correct about the relation of ergodicity and stationarity.

There is a spectrum of realization of a random process - it is a random function (for SB it is defined). There is also an energy spectrum, which is obtained by Fourier transform from ACF for stationary process (generalized to quasi-stationary). SB does not have an energy spectrum because it is neither stationary nor quasi-stationary. When they say spectrum of SB (they also say Brownian or brown noise), they mean a "tweaked" version of SB, which is a stationary process.

Um... Have you tried putting together a two-way auction equation? I have a suspicion that everyone here is doing the wrong thing. It's kind of like seeing to the root of the matter. Just small talk.