The Sultonov system indicator - page 72

 

I guess the model (or the primitive Y4AE↩calculus) looks like this;

for n=4.

And the check boils down to a comparison with x(i-th).

Right?

 
Nikolai Semko:
Yeah...
Interesting comrades you attract, Yusuf. :))

BLEEP, with your posts it's you " You attract..."

I'm sorry, "Yeah..."

No kidding.

 
Roman Shiredchenko:

BLEEP, with your posts you're the " You're pulling the..."

Sorry "Yeah..."

No kidding.

no worries

no offence... )))

to each his own

 
Сергей Таболин:

Calculation does not take part in the calculation. This is just a check of the formula.

Are all coefficients recalculated in the next calculation step?

Or is only a0 calculated, and the others are shifted and old a0 becomes a1, a1 becomes a2, etc.?

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

ps. If all the coefficients are recalculated it is a waste of time. This "indicator" only increases the uncertainty.(

If only a0 is recalculated and the check succeeds, this indicator deserves attention.

Imho.

 

So, a young man goes to the Internet, comes across this picture, and everything is serious: "mining and metallurgical institute"... but where is the scientific title, degree, position?

It reads:

"Consider a linear relationship"... Okay, linear is linear. If it's linear, what's the MNC? Where did the X's in quarts come from?

At the beginning it's a scheme, and then it all comes down to solving systems of equations.

What is interesting, earlier the author of this work wrote that he was from a commercial institute, but now it turns out that he is from Mining and Metallurgy Institute, but his degree, rank and position are unknown.

At our institute there was one charismatic guy who worked as a janitor in a school building, and at his leisure time he wrote erotic stories on the wall of the toilet... I must say he was quite talented - there is something to remember)))

 
Dmitry Fedoseev:

"consider a linear relationship"... OK, linear is linear. If it's linear, what's the MNC? Where did the X's in quarts come from?

You start out with a swing at something, and then it all comes down to methods for solving systems of equations.

Here is a mess of (at least) two problems. In the article, the problem posed with parameter estimation (or as it turns out fashionable to say nowadays - model learning)) is usually solved by ANC.
As for the topic, there is a somewhat different problem, the formal description of which we are trying to do correctly all together).

 
Dmitry Fedoseev:

So, a young man goes to the Internet, comes across this picture, and everything is serious: "mining and metallurgical institute"... but where is the scientific title, degree, position?

It reads:

"Consider a linear relationship"... Okay, linear is linear. If it's linear, what's the MNC? Where did the X's in quarts come from?

At the beginning it's a scheme, and then it all comes down to solving systems of equations.

What is interesting, earlier the author of this work wrote that he was from a commercial institute, but now it turns out that he is from Mining and Metallurgy Institute, but his degree, rank and position are unknown.

At our institute there was one charismatic guy who worked as a janitor in a school building, and at his leisure, he wrote erotic stories on the wall of the toilet... I must say he wrote quite brilliantly - there is something to remember)))

Gentlemen, it's not the answer, it's for everyone - it's past the "mining and metallurgical institute" - it's all good here - not this one... :-)

Here we have Altai Polytechnic and practice and work... :-)

 
Nikolai Semko:

no worries

I'm not offended... )))

to each his own

not at all

Glad :-)

Ledem das zein.

Sorry about the lyrics, not on purpose.

After P.S. You forced me to.

 
Mikhail Dovbakh:

Probably, the model (or the primitive Y-calculus) looks like this;

for n=4.

And the check boils down to a comparison with x(i-th).

Right?

It's more like this:

Y( i+k+1) =x(i+k+1) =a0 + aixi + ......+ a(i+k) x(i+k) , stressing that, Y(i+k) also refers to the original data set x with index (i+k+1), that is, the most recent value of x

 
Yousufkhodja Sultonov:

Rather like this:

Y(i+k+1) =x(i+k+1) =a0 + aixi + ......+ a(i+k) x(i+k) , stressing that, Y(i+k) also refers to the original data array xi

Dear Yusuf!

Don't introduce unnecessary (yet) entities. Hteor or Cteor. I agree. but where does Y come from? it's like from another dimension)))

However, the model (as you put it) is wrong again, a0 is fixed and does not account for i (

Also, I stick to the notation where the most recent point has a zero index).