Machine learning in trading: theory, models, practice and algo-trading - page 3735
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
The markup should be self-similar.
The geometry of space is very serious. But it is the one that gives an understanding of what is going on. But for that you need to know what space is, what geometry is. Then you need to know what symmetry is, etc. Any desire?
Artificial fractals are self-similar, natural fractals (including price fractals) are not self-similar.
That's a bit self-similar.
I was contrasting coordinates with force values. That is, when 0.9 is not more than 0.1.
And in standard approaches of feeding neural networks - 0.9 is more than 0.1.
This is a fundamental factor that determines the unreasonableness of the standard approach, because 0.9 of an indicator is a position, not a quantity.
That is, the price chart is vectors. A set of vectors. It is not a set of numbers.
And the models (most of them probably) have adders, which means that they will react to the value of a number. And most of them start with a graph. Then this value is "ironed" several times, and this "ironing" has nothing to do with some "profitable" statistics of this value.
This is mathematics for the sake of mathematics. That is, when all the power of the apparatus is used "in the wrong place". When they confuse tasks and fields of application.
Therefore, even if we set aside knowledge mining and learning without teachers, to teach models "something out there", and not specifically "something", then at least the input should be signal attributes, not quantitative ones.
Ivan Butko #:
Это уже отсебятина.
I apologise, of course, but if you don't know something, that's no reason to accuse everyone around you of making things up.
Ivan Butko #:
А в моделях (в большинстве наверное) есть сумматоры, а значит они будут реагировать на величину числа.
Imagine that there has long been a way to make it so that the models don't care about the magnitude of the number, and the only thing that matters is the fact that the numbers are different.
This is already an off-topic.
I was contrasting coordinates with force values. That is, when 0.9 is not greater than 0.1.
And in standard approaches of feeding neural networks - 0.9 is greater than 0.1.
This is a fundamental factor that determines the unreasonableness of the standard approach, because 0.9 of the indicator is a position, not a quantity.
That is, the price chart is vectors. A set of vectors. It is not a set of numbers.
And the models (most of them probably) have adders, which means that they will react to the value of a number. And most of them start with a graph. Then this value is "ironed" several times, and this "ironing" has nothing to do with some "profitable" statistics of this value.
This is mathematics for the sake of mathematics. That is, when all the power of the apparatus is used "in the wrong place". When they confuse tasks and areas of application.
Therefore, even if we set aside the extraction of knowledge and learning without teachers, to teach models "something there", not specifically "something", about at least the input should be signal signs, not quantitative.
Price movement on a chart is the resultant vector of the summation of the acting forces - bulls and bears. Some indicator shows the time derivatives of this movement. 0.9 velocity is greater than 0.1 velocity. In this case, velocity is a scalar quantity because we have a one-dimensional movement - up down.
I have a suggestion to go beyond the Brief Course of History of the All-Union Communist Party of Bolsheviks (Bolsheviks) of the Ministry of Defence, and to go deeper into the study of the structures of the SOURCE data (what you call your own rubbish).
Marginalised people are always dissatisfied with the current state of affairs and start inventing their own theories, which, when confronted with reality, turn them back to conventional concepts. This is the samurai's way, senseless and ruthless. Instead of learning from experience, they invent a bicycle.
If anything, Galileo was also a marginalist (in your terminology).
Who's good?
I'm good ;)