From theory to practice - page 357
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
There's a dialogue between me and A_K2 about Kolmogorov, and I went a bit wild today, too. It's not easy to get them out of hypnosis).
There's a dialogue between me and A_K2 about Kolmogorov, and I went a bit wild today, too. It's not easy to get them out of hypnosis.)
You do the math.
0.01 * 0.01 = 0.0001 - that's on the tails.
5 * 5 = 25 - that's in the centre.
what's the distribution? - hgngnh
Let them get a normal one, otherwise they ' re already shying away from science.
//not a word about physics and Schrodinger...
and then we'll listen to the senior scientists, of course...
Think about it.
0.1 * 0.1 = 0.01 - that's on the tails.
5 * 5 = 25 - it is in the centre.
What's the distribution? - I don't know.
Let them get the normal one, because they are not ashamed to have the grails and science.
//Not to mention physics and Schrodinger...What for? For any forecasting, there are no requirements for the type of distributions at all. The requirements for stationarity are only within the framework of the article, not in general.
Forum on trading, automated trading systems and trading strategies testing
From Theory to Practice
Yuriy Asaulenko, 2018.05.08 15:09
It doesn't matter at all. Already in the notorious article by Kolmogorov, 1940, there are no requirements for the form of distribution at all. There is only one requirement for the series there - stationarity. And the article considers the possibility and forecasting conditions only and exclusively for the stationary time series.
Where did the normality requirement come from in the topic? - I do not know)).
HZ the possibility of predicting non-stationary series in the article simply is not considered, ie, the article in relation to such series simply says nothing. It absolutely does not mean impossibility of forecasting of nonstationary series, but simply, in relation to nonstationary series, means nothing at all.
I ask respected traders to reread Kolmogorov once again. Without emotion.
Please note that this is the only work by a world-renowned scientist dedicated to forecasting. There are no other methods in nature from people equal to Kolmogorov in intelligence.
It's tempting to put it all into practice.
Here I have to agree with Yury Asaulenko - there is not a word about the normal distribution. Only stationarity with additional conditions to the correlation function.
Is it possible to transform BP to stationary form by Erlang flows? No one has proved the contrary.
Why by Erlang flows? WITH WHAT ELSE?!?
Well, I just reread Feynman again. He mentioned in passing that it would be good to predict market prices. And then immediately started giving analogies with events - raindrops, Geiger counter, etc. Considered tau=Time/number of events. Concluded that the best model is a Poisson flow of events. Then covered everything with probability amplitudes. Voila - the dreaded triple integrals describe the probability of this or that event.
Further he recommends to work on "rather large" times. He doesn't explain how big they are.
If this undertaking with Erlang's flows will not give me anything - I will wash my hands of it. But it must be brought to its logical conclusion.
I ask respected traders to reread Kolmogorov once again. Without emotion.
Please note that this is the only work by a world-renowned scientist devoted to forecasting. There are no other methods in nature from people equal to Kolmogorov in intelligence.
It's tempting to put it all into practice.
Here I have to agree with Yury Asaulenko - there is not a word about the normal distribution. Only stationarity with additional conditions to the correlation function.
Is it possible to transform BP to stationary form by Erlang flows? No one has proved the contrary.
Why by Erlang flows? WITH WHAT ELSE?!?
Well, I just reread Feynman again. He mentioned in passing that it would be good to predict market prices. And then immediately started giving analogies with events - raindrops, Geiger counter, etc. Considered tau=Time/number of events. Concluded that the best model is a Poisson flow of events. Then covered everything with probability amplitudes. Voila - the dreaded triple integrals describe the probability of this or that event.
Further he recommends to work on "rather large" times. He doesn't explain how big they are.
If this undertaking with Erlang's flows will not give me anything - I will wash my hands of it. But it must reach its logical conclusion.
Forum on trading, automated trading systems and testing of trading strategies
The Zigzag Indicator and Neural Networks
Prival, 2007.12.01 23:54
Here I do not agree with you. There are these works, only they should be applied to the Forex market.
"Machines won't learn to think until humans learn to think"
The first fundamental results in the theory of filtering in discrete time were obtained by the Soviet scientist A.N. Kolmogorov [1] (1941), and in continuous time - by American scientist N. Wiener [2] (1942). Complete results on linear filtration theory of Gaussian processes in discrete and continuous time were obtained by R.E.Kalman and R.S.Bucy [3] (1960, 1961). Fundamental results in nonlinear filtering theory belong to Soviet scientist G.L.Stratonovich who worked out the theory of nonlinear filtering of Markov random processes since 1959 [4,5,6, etc.].
There are also works by Levin B.R. http://www.computer-museum.ru/connect/levin.htm and Tikhonov V.I.
I would give all Nobel prizes to Stratonovich for his GREAT equation. I just think one should be able to prepare it (the equation) for Forex. Which is what I am trying to do now.
I ask respected traders to reread Kolmogorov once again. Without emotion.
Please note that this is the only work by a world-renowned scientist devoted to forecasting. There are no other methods in nature from people equal to Kolmogorov's intelligence.
Well, well. No, there isn't. But there have been many works on prognosis since the early 40s of the 20th century. I can't name the authors yet, but they are all well known. And both stationary and non-stationary processes were considered. Why do you think they have all gone mad, and all at the same time?
Is it possible to transform BP to stationary form by Erlang flows? The opposite has not been proven by anyone.
Why Erlang flows? WHAT ELSE?!?
....
If this Erlang flows thing doesn't work, I'll wash my hands of it. But, it has to be brought to its logical conclusion.
It won't. Just write it down somewhere for your memory.)
I ask respected traders to reread Kolmogorov once again. Without emotion.
Please note that this is the only work by a world-renowned scientist dedicated to forecasting. There are no other methods in nature from people equal to Kolmogorov in intelligence.
It's tempting to put it all into practice.
Your notorious Kolmogorov describes a common autoregressive model.
He may have been a pioneer, but by now this class of models has been embodied a million times in every conceivable form. It is described in any textbook on series analysis. Google is the help.
SanSanych wrote about it somewhere in the middle of the thread. But that's not the point.
Your notorious Kolmogorov describes an ordinary autoregressive model.
He may have been a pioneer, but by now this class of models has been embodied in every conceivable form. It is described in any textbook on series analysis. Google is the help.
SanSanych wrote about it somewhere in the middle of the thread. But it is not the point.
So be it. Other forecasting methods I have not found, alas... (The task was to find the work of a truly outstanding person).
But, it turns out one single thing - only bringing BP to a stationary form can give results in neural networks. That's what I'm getting at. There is no other way and there cannot be. Right?
PS For general understanding of the current moment - TC based on diffusion equations has already been created, works and is no longer of interest to me. I want to switch to neural networks. Maybe this change of general reasoning in this branch causes misunderstanding?
Your notorious Kolmogorov describes an ordinary autoregressive model.
He may have been a pioneer, but by now this class of models has been embodied a million times in every conceivable form. It is described in any textbook on series analysis. Google is the help.
SanSanych wrote about it somewhere in the middle of the thread. But it does not help.
))) Fail. The fag had sneaked up on me, though he was visible from afar.
* Squeak is a fat polar fox *
But, it turns out one thing - only bringing BP to a stationary form can produce results in neural networks. That's what I'm getting at. There is no other way and there cannot be. Right?
Why? NS generally don't care about your stationarity at all. I already wrote yesterday about what the NS needs. I realised you didn't get it. It's not the point.