Machine learning in trading: theory, models, practice and algo-trading - page 2636

New comment
 
Maxim Kuznetsov #:

These "theorists". :-)

There is nothing practical about futile attempts to multiply entities beyond what is necessary. This is the third area of activity - not theory or practice, but chattering away.

There is nothing but volatility, directional movement and looking for opportunities to highlight the latter amidst the fluctuations of the former.

 
as you like
 
Aleksey Nikolayev #:

There is nothing practical about futile attempts to multiply entities beyond what is necessary. This third area of activity is neither theory nor practice, but empty chatter.

There is nothing but volatility, directional movement and looking for opportunities to highlight the latter amidst the fluctuations of the former.

In markets there is abstract time as well as abstract price and consequentlyabstract regularities.

This is difficult to perceive in a two-dimensional understanding of the world on the chart. You have to look more broadly.

 
Aleksey Nikolayev #:

There is nothing practical about futile attempts to multiply entities beyond what is necessary.

Hasn't that been going on in this thread for a year now?)
 

Is it really that difficult to make 2% a day or even an hour by hand?

Or is it more profitable to build a money printing machine for decades, sorry TCSMO with 2% a year.

Time is money after all).

 
secret #:
Hasn't that been going on in this thread for a year now?)

Because of the lack of meaningful moderation, there are plenty of entities who do not know the topic (as outlined in the very first post of the thread), but want to saysomething. You and your compatriot have made your mark.

 
Aleksey Nikolayev #:

Because of the lack of meaningful moderation, there are plenty of entities who do not know the topic (as outlined in the very first post of the thread), but want to saysomething. You and your compatriot have made your mark.

Good luck on this endless journey).
 
Think of features like increments, but more informative. For example, find the average price for the entire history and deduct the rest from it. You want the scatter to be as large as possible, but within a range known from the new data.

Fractional differentiation works this way (maximum scatter when stationarity is maintained), but I want something new.

Maybe some "slope lines" from time and subtract from them prices, decibels, f-from time, any kind of turbidity, as long as the stationarity and maximum dispersion conditions are met.
 

Suppose we have found patterns that occur periodically and accompany a particular price movement once they occur.

Has anyone done any research on the relationship between the frequency of occurrence of a pattern and the subsequent event?

We are talking about probability clusters, if there is such a term.

Suppose we can expect that if a pattern has not appeared for a long time, there will be a predictable (concomitant) price movement after it has appeared, and then there will be a fading as the pattern has become visible to all and thus eliminated market inefficiencies.

I think that the development of metrics to assess these transient states over time (from more likely to equally likely or even negative prediction) may help to find and select such patterns, and a model that can account for this may prove quite effective.

I am working in this direction, but I lack the mathematical apparatus and theoretical knowledge.

 
secret #:
Good luck on this endless journey)

Cheerio to you too, and long walk on the path you know)

1...262926302631263226332634263526362637263826392640264126422643...3522
New comment
Reason: