The Sultonov Regression Model (SRM) - claiming to be a mathematical model of the market. - page 29

 
avatara:

I see no reason to deny Yusuf's innovation and named regression models.

It's not like anyone is being dragged by Bulashov's forks and Elliott waves...

Everyone invents to the extent of their knowledge.

And the famous (18) is better than some alligators. )

Incomprehensible (set, by the way, by Alexey - mathematician) disparaging tone of discussion of one of mathematical models surprises - this is a specialized forum. There is a strange childish jealousy. Like I myself have not invented anything, so it is up to others. Even Yusuf has a complex - he took a penny for his patent in the USSR, but if he had done so now or then abroad he would have made a good living. That is why he is immediately going to work on land))).

Moreover, the joker - attributes to his model miraculous properties almost from all diseases and the evil eye...

;)

I would suggest that fans of extrapolations move on to constructive discussion, and the verbally-minded cozy talkers go to the smoking room.


+100
 
Integer:
There is no invention here, not even the solution to an engineering problem, the solution to which usually leads to an invention. Here is simply the solution to a mathematical problem. Which is not complex and unsolvable, like the problem solved by Perelman. Ordinary routine mathematics.

Yeah. And Chuvashov's spoon?

:)

Another annoying thing is the disregard for the title of associate professor.

It's not enough to be a Hodge, but also a docent. plus age.

As a schoolboy with glowing eyes - bad. docent - bad.

Militant ignorant Mishek - good ((

 

HideYourRichess 12.07.2012 13:08

yosuf:
Linear regression applies when you assume that there is a linear relationship between price and time,
A point of principle. Price doesn't depend on time, the price changes with time, but it doesn't depend on it. What the price really depends on is the willingness to buy or sell.
Mischek2 12.07.2012 13:44
avtomat:

You, pardon me, have uttered utter nonsense...

;)))


And it seems to me that you

Or does TIME move the price ?

------------------------------------------------------------------------------------

We are dealing with time series with memory, i.e. subsequent values depend on previous values. Take any regression equation in econometrics - they are all functions of time.

 
avtomat:

And you seem to find it difficult to see the interconnections deeper than one level...

Yeah, the tail wags the dog
 
faa1947:

HideYourRichess 12.07.2012 13:08

yosuf:
Linear regression applies when you assume the existence of a linear dependence of price on time,

A point of principle. Price doesn't depend on time, price changes with time, but it doesn't depend on time.

That's what the price really depends on, it's the willingness to buy or sell.

Mischek2 12.07.2012 13:44
avtomat:

You, pardon me, have uttered utter nonsense...

;)))


And in my opinion you

Or does TIME move the price ?

------------------------------------------------------------------------------------

We are dealing with time series with memory, i.e. subsequent values depend on previous values.

Take any regression equation in econometrics - they are all functions of time.

I'm not clear on your position - if the time function is independent of time, what kind of statement is that?

:)

 
avatara:

Yeah. And Chuvashov's spoon?

:)


You yourself know how the community reacts to that phrase. And the spoon itself isn't that bad.
 
Mischek2:
Yeah, tail wagging the dog
that's the real deal -- a little bit of a pickle in the garden...
 
avatara:

I see no reason to deny Yusuf's innovation and named regression models.

It's not like anyone is being dragged by Bulashov's forks and Elliott waves...

Everyone invents to the extent of their knowledge.

And the famous (18) is better than some alligators. )

Incomprehensible (set, by the way, by Alexey - mathematician) dismissive tone of discussion of one of mathematical models surprises - this is a specialized forum. There is a strange childish jealousy. Like I myself have not invented anything, so it is up to others. Yusuf has a complex - in the USSR he earned a penny for his patent, but if he had done so now or then abroad he would have made a good living. That is why he is immediately going to work on land plots )))

He is also a funny guy - he ascribes to his model miraculous powers against almost all diseases and the evil eye...

;)

I would suggest that those who like extrapolations move on to constructive discussion, and the verbally inclined waddlers go to the smoking room.


I have tried several times to get constructive from Yusuf, including this thread.

I wish the econometrics teacher would use econometrics terminology in a standard way. One last thing: the linearity or non-linearity of a regression is determined by the kind of parameters of that regression, not by the kind of independent variables (arguments of the function), which can be anything. But the type of parameters strictly requires appropriate methods of their estimation, the most common MOC is applicable to linear parameters, i.e. linear regressions, and on top of that under a number of rather strict restrictions.

My guess (???) is that using a gamma function allows for smoothing that remains differentiable to the right at the arrival of a new bar. Virtually all smoothing methods do not have this property. There is no way to find out.

 
avatara:

your position is not clear - if the function of time is independent of time, what kind of statement is that?

:)

Read again the two lines of my post above yours. If you don't understand it after the 10th time, switch to reading one line posts.
 
faa1947:

I have tried several times to get constructive feedback from Yusuf, including this thread.

I wish the econometrics teacher would use econometrics terminology in a standard way. One last thing: the linearity or non-linearity of a regression is determined by the type of parameters of that regression, not by the type of independent variables (arguments of the function), which can be anything. But the type of parameters strictly requires appropriate methods of their estimation, the most common MOC is applicable to linear parameters, i.e. linear regressions, and on top of that under a number of rather strict restrictions.

My guess (???) is that using a gamma function allows for smoothing that remains differentiable to the right at the arrival of a new bar. Virtually all smoothing methods do not have this property. There is no way to find out.

Yusuf's innovation is that he abstracted from the a priori description of the model by plausible functions.

Have you looked at pin 18?

There's an interesting twist there...

;)

Reason: