The Sultonov system indicator - page 116

New comment
 
Yousufkhodja Sultonov:

Renat, I never said that. I said that I wouldn't judge until I'd tested everything on a real account. I'm waiting for the advisor to transfer from MKL5 code to 4.

ok

debates

---

I would personally suggest asking this person's opinion - is it worth it?

For he shoots a tenner, and only engages in things that are worthy of attention
 
Yousufkhodja Sultonov:

Bravo, you have outperformed Gauss and Kramer:

Consider the linear dependence of the exponent Y on a set of variables x:


To estimate the coefficients of the equation, we apply Gauss's method of least squares and obtain the following system of k linear equations with at least n ≥ k+1 groups of actual data Y depending on the values of variables x:


In general, this system of equations is solved by Gauss (1777- 1855) method of successive elimination of variables or by using the properties of matrices, known as Cramer's method (1704-1752).

Computational complexity

The method of Gauss is a classical method for solvingsystems of linear algebraic equations(SLAE). This is a method of sequential elimination ofvariables, when using the elementary transformation of a system of equations is reduced to an equivalent system of a stepped (or triangular) form, from which consistently, starting with the last (by number) of variables, are all the remaining variables.

The algorithm of solution ofSLAE by Gauss method is divided into two stages.

  • At the first stage the so-called direct course is carried out when byelementary transformations over strings the system is reduced to the stepped ortriangular form, or it is established that the system is incompatible. Namely, among the elements of the first column of the matrix one chooses a non-zero one, moves it to the extreme upper position by permutation of rows and subtracts the first row obtained after permutation from other rows, having multiplied it by the value equal to the ratio of the first element of each of these rows to the first element of the first row, thus zeroing out the column under it. After the above transformations have been made, the first row and the first column are mentally crossed out and continued until a matrix of zero size remains. If at any iteration no nonzero element is found among the elements of the first column, one proceeds to the next column and performs the same operation.
  • At the second stage the so called backward procedure is carried out the essence of which consists in expressing all obtained basic variables through non-basic ones and construct a fundamentalsystem of solutions or, if all variables are basic, express numerically a single solution of the system of linear equations. This procedure starts with the last equation, from which the corresponding basis variable is expressed (and there is only one) and substituted into the previous equations, and so on, going up in "steps". Each line corresponds to exactly one basis variable, so at each step except the last (topmost) one, the situation exactly repeats the case of the last line.

Cramer's methodrequires computation ofdeterminants of appropriate dimension. When usingthe Gaussian methodto compute the determinants, the method has a time complexity of order4, which is worse than ifthe Gaussian methodwere directly used to solve the system of equations.

Yusuf, what would you say to a student who would come to you for a term paper and inside it?

Where and how do you apply Gauss ISC? What do you call the "estimation" of the coefficients? And how is this "next system" derived?

Why such big quotes about the elementary? First like swing, one two, back and forth, and the quote is about solving a linear system of equations. Who is this meant for? Total idiots?

 
Олег avtomat:

Get a matcad -- do you show a lot of cleverness?

Have you demonstrated much? You demonstrate school problems solvable by one line, and five screens of picturesque canvases. And those aren't always correct.

 

It seems to be outdated methods, SVD is used by all))

but for tajikistan this will do

 
Maxim Dmitrievsky:
It seems to be outdated methods, SVD is used by all ))

What is SVD (Dragunov sniper rifle?), speak in human language, please. Reminds me of a teacher who also taught TAU, by the way... He was so smart, too, all like: "you know this from maths..." - opens up a textbook and copies the formulas one-to-one from the textbook. It's the same here. I have a lot of questions about these associate professors at universities - what do they know and understand and whether they themselves know and understand it? It's the same here with Yusuf - "Everything here is so simple, one, two" - and it was rewritten from some textbook, and it seems that it was rewritten completely wrong.

 
Dmitry Fedoseev:

What is SVD, speak in human language, please. Reminds me of a teacher who also taught TAU... He was also so smart, all like: "you know this from maths," opens up a textbook and copies the formulas one-to-one from the textbook. The same thing is happening here. I now have a lot of questions about these associate professors from institutions - what do they know and understand and whether they themselves know and understand it? Same here with Yusuf - "here everything is so simple, one, two" - and rewritten from some textbook, and it seems to be completely wrong rewritten.

singular matrix decomposition

 
Maxim Dmitrievsky:

singular matrix decomposition

So it's recognisable? Plucked from some textbook without understanding and passed off as a masterpiece of your own thought?

 
Dmitry Fedoseev:

You mean recognisably? Taken from some textbook without understanding and passed off as a masterpiece of your own thought?

What's recognizable? It's just translated.

 
Maxim Dmitrievsky:

What's recognisable?

What is SVD then?

 
Dmitry Fedoseev:

What is SVD then?

singular matrix decomposition, or just singular decomposition

google is working today.

http://www.igce.comcor.ru/AI_mag/svdexam.html

1...109110111112113114115116117118119
New comment