Principles of working with an optimiser and basic ways of avoiding fitting in. - page 7

New comment
 
Mathemat:

I wonder from what principles it follows that it should be reversible (in time, or what?)...

Even physicists have known for decades that there is no perfect symmetry in nature.

Formulated in the post, without philosophy.
 
faa1947: Formulated in the post, no philosophy.
It is not an answer. Philosophy is assumed, right?
 
Mathemat:
That's not the answer. Philosophy is assumed, right?
You add up the right side and you get the left side. That's it.
 
Somewhere went the wrong way...))
 
faa1947: Add up the right side and you get the left side. That's it.
No way. Why the hell do we have to add up the future to get the past?
 
LeoV:
They went somewhere in the wrong direction...)))


But with clever faces on their faces

And they were talking nonsense (c) Y. Moritz

 
LeoV:
Going somewhere in the wrong steppe...)))

There are principles which are not proved but discussed. The author of the thread is sticking to the "history repeats itself" principle of TA. It is generally accepted.

I sit in a different value system which also has the principle of reversibility of the model. In this thread this principle is fundamental. If our TS takes only a part of information from a quotient, for example the trend, won't it turn out that what has been discarded is more important than what has been taken? For me, and not only for me, this question is crucial. In the frame of the topic: without observing the principle of reversibility, we test a part of information from the quotient and according to the results of the test, not taking into account the discarded part, we try to extrapolate the results of the test into the future.

So that steppe, still native.

 
Mathemat:
Nizachod. Why the hell do we have to fold the future to get the past?
What does the past have to do with the future. quote = 1.3215 decomposed into two components 1.3200 and called the trend, and the rest called the noise. Maestro, are you stupid or what?
 
faa1947:

The author of the thread is sticking to the "history repeats itself" principle of TA. It is generally accepted.

It is. Only it doesn't repeat itself the way you think it does. Gg :)))
 
faa1947: What has the past, the future got to do with it. quote = 1.3215 decomposed into two components 1.3200 and called a trend, and the rest is called noise. Maestro, are you stupid or what?

You were talking about a different kind of reversibility. I'm not arguing about the one you're talking about now. But it was like that:

faa1947: But there's another add-on: the reversibility of the model is the quotient on the left, and everything on the right should sum to that quotient.

Translated into Russian: "Everything that we calculate on the right (in the future) must [after the proper transformations] give what is on the left (the past)". It turns out to be a sort of faa model verification.

Am I still dumb? Or am I just playing with terms, because I got bored?

1234567891011
New comment