How do you properly compare two non-overlapping rows? - page 8

 
transcendreamer:

I wonder, what is the point of the professor trying to convince that hehas invented a non-linear ISC when clearly this is not the case?

Some kind of hypertrophied hubris or something...

Or has the professor not heard of the Levenberg-Marquardt method/algorithm?

And that's the middle of the last century for a second.

The good news is that the algorithm is already in the ALGLIB library, i.e. you don't have to write anything yourself.

There is both documentation and open source codehttps://www.alglib.net/optimization/levenbergmarquardt.php.

Everything has been optimized there for a long time for speed and accuracy (L-BFGS).

There are ready-made lsfitfit functions - you can just take it and use it.

The main thing is that it is available in MT5 because alglib was ported natively to mql5 long time ago.

The only disadvantage is that you have to know better than school mathematics and be able to work with barycentric coordinates and hessians.

But in general, you can use polynomials or splines, with or without weights or with constants...

Why reinvent the wheel with such a strain?

Let alone that ISC is not the unique approach and other approaches (loss function, maximum likelihood, etc) have been known for a long time.

Again this has long been known since the 60's at least.

You can read it here:http://statsoft.ru/home/textbook/modules/stnonlin.html

Simply, you haven't tried to get into the essence of PNB functions, leaving behind all your reasoning about ISC of all its varieties, including the link you cited. In this situation, I would not argue with a "professor" as you put it, although I am only an acting associate professor.

 
Yousufkhodja Sultonov:

Simply, you have not tried to get into the essence of PNB functions, leaving behind all your arguments about the ISC of all its varieties, including the link you cited. In this situation, I wouldn't be arguing with a "professor" as you put it, although I am only an acting associate professor.

Please do not be offended, I was just stating my personal opinion.

Is there anyone at all who has gone into the essence of PNBs and started applying them? 😉

 
transcendreamer:

Please don't be offended, I was just expressing my private opinion.

Is there anyone at all who has got into the essence of NSPs and started applying them? 😉

There is no one so far. Everyone thinks it's too complicated and cumbersome, but it's not.

 
Yousufkhodja Sultonov:

There is no one yet. Everyone thinks it's too complicated and cumbersome, but it's not.

No, it's not. It's like the Elusive Joe joke.

 
Yousufkhodja Sultonov:

There is no one yet. It seems to everyone that it is too complicated and cumbersome, although this is not the case.

If the method is as similar as possible to others that have long been known and described in detail, and if the "experiment" does not yield any results, then it is simply not clear what and why pay attention to it? Your attempts to draw attention in the absence of any outstanding results are either the result of a bluntness that does not allow you to understand the fact that you have invented a bicycle, or a simple-minded insolence.

 
vladavd:

Your attempts to attract attention in the absence of any outstanding results are either from an obscurity that prevents you from recognising that you have invented the bicycle, or from unsophisticated unpretentious impertinence.

The Doctor's patients are difficult to understand. Only the standard one comes to mind - the psychiatrist's main mistake is to understand how the patient thinks.

 

I had a professor at uni who tried to patent the Wheatstone bridge. He was talking in earnest about how good he was and why what he was doing was incredibly cool. It's the kind of thing that's used in voltage measurements.

And he genuinely didn't understand why he was rejected. It is true that he made some kind of change, I can't remember which one, but it was so insignificant that any technically competent person would have understood it perfectly well.

 
Maxim Romanov:

I had a professor at uni who tried to patent the Wheatstone bridge. He was talking in earnest about how good he was and why what he was doing was incredibly cool. It's the kind of thing that's used in voltage measurements.

And he genuinely didn't understand why he was rejected. It is true that he made some changes, I do not remember which ones, but they were so insignificant that any technically literate person would have understood them.

And we had an inventor of the planetary gearbox. In short, a topic for study by the relevant specialists).

 
Yousufkhodja Sultonov:

Simply, you have not tried to get into the essence of PNB functions, leaving behind all your arguments about the ISC of all its varieties, including the link you cited. In this situation, I would not argue with a "professor", as you put it, although I am only an acting associate professor.

Communicating with quite a lot of PhDs I noticed one interesting thing, those who did real things that worked really well in reality, could in a few simple sentences explain the essence of work and principles why it works.
You are not capable to explain the essence of work at the associate professor level, only at the level of the usual forum "guru" who supposedly knows and understands everything, and he always has a "brilliant" idea in his pocket for all occasions.

You know, one clever fellow here said a beautiful phrase:
"When you have a hammer in your hand, everything seems like nails" and I completely agree with him.

You do not have to answer me, as I do not want to argue with you.
Just draw conclusions, if you are capable of doing so at all.
 
You need to combine the Theil-Sen method with a planetary gearbox and a Wheatstone bridge, then the PNB will work.
Reason: