Machine learning in trading: theory, models, practice and algo-trading - page 2556

New comment
 
Aleksey Nikolayev #:

Vorontsov is probably the best expert on MO in Russia. The course is therefore bound to be good, but since it is for IT people, it omits basic and important mathematics for us. I`ve noticed more than once, that for the application of mathematical methods in trading their basic, simplified form is not very suitable.

MO is based (see for example Tibshirani) on the assumption that there is a constant joint distribution of predictors and responses P(X,Y). From it, the conditional probability Py(Y|X) can be calculated, from which the regression Y=f(X) can be calculated. Eventually, this regression is approximated by some MO models. In the physical world, this theory more or less works. But not in trading. It turns out that P(X,Y) changes unpredictably with time (non-stationarity) and the whole theory collapses a bit.

The most popular approach is just to ignore non-stationarity and then be surprised with the results and complain about the MO).

Well, there is an interesting part in the second part, at the end, about the time series and his experience with it. The rest is up to everyone
Non-stationarity is not as critical as lack of regularity. If we assume that the time series is unpredictable at all, I'm afraid there's nothing more to be invented here
 
Aleksey Nikolayev #:

MO is based (see for example Tibshirani) on the assumption that there is a constant joint distribution of predictors and responses P(X,Y). From it, a conditional probability Py(Y|X) can be calculated, from which a regression Y=f(X) can be calculated. Eventually, this regression is approximated by some MO models. In the physical world, this theory more or less works. But not in trading. It turns out that P(X,Y) changes unpredictably with time (non-stationarity) and the whole theory collapses a bit.

The most popular approach is just to ignore non-stationarity and then be surprised with the results and complain about the MO).

You couldn't have said it better.

Well done, but what to do?

 
Maxim Dmitrievsky #:
Non-stationarity is not as critical as lack of regularity.

How is regularity measured?

 
mytarmailS #:

How is regularity measured?

correlation, entropy

Maybe there are others.

 
Maxim Dmitrievsky #:

correlation, entropy

Maybe there are others.

What do you mean? Correlation, entropy.

What with what, when, why?

On the Internet the definition of irregularity is when there are gaps in dates with observations, what else do you mean?

 
mytarmailS #:

Meaning? Correlation, entropy...

What with what, when, why?

On the internet, the definition of irregularity is when there are gaps in dates with observations, are you referring to something else?

cycles

 
Maxim Dmitrievsky #:

cycles

A straight line has no "regularity" or "cycles," but it is predictable. There are many examples of this

Non-stationarity is a problem.

 
Maxim Dmitrievsky #:

cycles

there are no cycles...

there may be complex sums of loops (interfection) but there are no loops in the usual sense

 
Dmytryi Nazarchuk #:

A straight line has no "regularity" or "cycles," but it is predictable. There are many such examples.

Non-stationarity problem.

The sloping straight line is nonstationary, and in fact we are talking about time series.

Stop talking nonsense, where did you weirdos come from again? :D it's just worth warming up the topic.

 
mytarmailS #:

there are no cycles...

there may be complex sums of cycles (interfection) but no cycles in the conventional sense

it's clear that the quotes are non-stationary and it's cycles that are searched for

1...254925502551255225532554255525562557255825592560256125622563...3647
New comment