The Sultonov Regression Model (SRM) - claiming to be a mathematical model of the market. - page 43
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
I will check, but in general you are wrong, since 0 is a discrete value, and you use a continuous normal distribution law, so you should introduce a generalized density, since the random variable is mixed X, with possible values of x, which takes one discrete value of 0, the rest continuous values!
There is a notion of a "discrete random variable" - which can take a countable (not necessarily finite) set of values (e.g. the number of eagle falls in a series of experiments). For such quantities a so-called probability distribution is defined - i.e. it is a set of probabilities of a value hitting certain points. If we consider it as a function, it will indeed be bounded by a segment from 0 to 1.
On the other hand, there are "continuous random variables", i.e. those whose set of possible values is continuous. They have a distribution function and a probability density function - and the first function is always non-decreasing, equal to 0 at minus infinity and 1 at plus infinity. The density of the distribution is its derivative, it can take any non-negative value, including being infinitely large at certain points, as long as the integral of it on the entire number axis is equal to 1. The density of a distribution is not a probability of anything, so there is no meaningful restriction on its values.
PS Once we all learn the terms, 90 per cent of the arguments will disappear from the forum.
PPS Yusuf, you are becoming increasingly sad to read (
alsu:
For orb:
PS As soon as we all learn the terms, 90 per cent of the arguments will disappear from the forum.
I would loosen the requirement: replace the word 'learned' with 'looked it up in a textbook'. Apart from that I would wish to refrain from words like "ignoramus" especially to people who can support their point of view with calculations in a complicated system of statistical analysis.
PS As soon as we all learn the terms, 90 per cent of the arguments will disappear from the forum.
I would loosen the requirement: replace the word "learned" with "looked it up in a textbook". Apart from that I would wish to refrain from words like "ignorant" especially to people who can support their point of view with calculations in a complex system of statistical analysis.
Description of the sine Y = Sin(0,1x)+2 using RMS:
1. Direct RMS:
2. reverse RMS:
3. Averaged RMS:
It's beautiful.
But my imho - there can be no mathematical model of the market at all.
Description of the sine Y = Sin(0,1x)+2 using RMS:
1. Direct RMS:
2. reverse RMS:
3. Averaged RMS:
Looking at these figures, it is safe to say that RMS is of no use. Either we enter RMS signals too late, or we get false signals. And this is on such simple functions as sine and cosine.
It's beautiful.
But my imho - there can be no mathematical model of the market at all.
Why not?
If we (hypothetically) rewrite all traders with all their capabilities and characteristics, including statistical ones, and the environment at the same time, there you have a market model. Of course, it's too cumbersome and therefore most likely unsuitable for practical use, not to mention the fact that in practice it certainly would not work. Still, there is no ban in principle on creating such a "comprehensive" model, which means that the model itself is perfectly valid. Rather, the problem is that we want to simplify it so much that it can be put into a desktop computer, and so that the computation time is acceptable.
In short, I believe this task is doable. And I even believe that someone sooner or later will be able to get a model. Although, perhaps, one should have no illusions: if a person is smart enough to describe the market with a simple enough model, he will be smart enough to keep quiet about it #_# (yes, this is a stone to Yusufkhoja)
Why not?
If we (hypothetically) rewrite all traders with all their capabilities and characteristics, including statistical ones, and the environment at the same time, here you have a market model. Of course, it's too cumbersome and therefore most likely unsuitable for practical use, not to mention the fact that in practice it certainly would not work. Still, there is no ban in principle on creating such a "comprehensive" model, which means that the model itself is perfectly valid. The problem is rather that we want to simplify it so much that it can be put into a desktop computer, and so that the calculation time is acceptable.
In short, I believe this task is doable. And I even believe that sooner or later someone will manage to get a model. Although, probably, don't have any illusions: if a person is smart enough to describe the market by a simple enough model, he will be smart enough to keep quiet about it #_# (yes, that's a stone to Yusufkhoja)
I have tried to express the average forecast price of a future bar (F) through OHLC prices of previous bars as following dependence, though I do not know if they have tried it before in such a form or not
F=A*O^a1*H^a2*L^a3*C^a4,
where - A, a1, a2,a3,a4 are constant coefficients, defined by Gauss ANC method and this is what I got for 15 bars of TF D1:
Hence, the quotient can, in principle, be expressed by a single equation, but let's find out what the practical usefulness of this is. What are your views?