Machine learning in trading: theory, models, practice and algo-trading - page 2404

[Aleksey Nikolayev]

Aleksey Nikolayev:

For example, some simple model retraining where one new example is added and obsolete ones are thrown out.

Vorontsov's presentation as an illustration of the idea of the approach.

[Maxim Dmitrievsky]

Aleksey Nikolayev:

Vorontsov's presentation as an illustration of the idea of the approach.

On-the-fly expert optimization using machine learning: logit regression

https://www.mql5.com/ru/articles/6351

[Maxim Dmitrievsky]

Or all sorts of one-armed bandits and reinforcement learning for time series that updates states over time

doesn't work in the market, but you hang in there.

[Aleksey Nikolayev]

Maxim Dmitrievsky:

Or all sorts of one-armed bandits and reinforcement learning for time series that updates states over time

Doesn't work in the market, but you hang in there.

Well, there's all sorts of stuff out there, including change point detection.

It's not hard to find something that doesn't work in the market - it's hard to find something that does work.)

[Renat Akhtyamov]

Aleksey Nikolayev:

Well, there's all sorts of stuff, including change point detection.

Anything that doesn't work on the market isn't hard to find - it's hard to find anything that works)

ha,

anything

there is only one option

;)

[Evgeni Gavrilovi]

Maxim Dmitrievsky:

reinforcement learning for time series that updates states over time does not work in the market

Then what are the most promising methods in MO?

[Maxim Dmitrievsky]

Evgeni Gavrilovi:

Then what are the most promising methods in MO?

it depends on what tasks, I think generative and contextualized, in general

they are constantly evolving, you never know what will happen next

[mytarmailS]

AHAHAHH....

I decided to recognize pictures with a stochastic, which MO method would be more promising? )))))

[Aleksey Nikolayev]

Aleksey Nikolayev:

Well, there's all sorts of stuff, including change point detection.

Meaning something like what is described in this article (the article itself is not very useful).

[Maxim Dmitrievsky]

Aleksey Nikolayev:

Meaning something like what is outlined in this article (the article itself is not particularly useful).

the idea is correct in general, but it does not necessarily require online training in real life, it can be done only at the stage of basic training / retraining, and then used as it is