Machine learning in trading: theory, models, practice and algo-trading - page 2051

New comment
 
Aleksey Vyazmikin:

And the approach itself is interesting, how many points were used to describe the structure?

20 forward 20 backward from the center

 
mytarmailS:

clustering is very good...

I just want to describe on the level of images what I can't describe by code myself...

It's simple...


the work of an automatic recognizer...

of course it doesn't work )) but I got what I wanted...

The idea is good.

How many patterns were found in total and how many of them are "working"?

Clustering has created many clusters, among which there are similar ones, as it seems to me.

mytarmailS:

20 forward 20 backward from the center

Did you click only the center point?

 
Aleksey Vyazmikin:

This is a good idea.

How many patterns were found in total, and how many of them are "working"?

Clustering created many clusters, among which there are similar ones, as it seemed to me.

Objective: I have a pattern that I seem to see, but I can't describe it in code


I look at the chart and click those points that I think are in the center of this pattern.

2 I collect a database of these clicks

3 I cluster what I clicked on.

4 I select from the prototypical clusters what, in my opinion, is the "averaged image" of the pattern I am looking for.

5. I drill down on this image and go through the graph and compare the current data with the image for similarity

Aleksey Vyazmikin:

Did you click only the center point?

YES
 
mytarmailS:

Problem: I have a pattern that I seem to see, but I can't describe it in code


1 I look at the chart and click on what I think is the center of this pattern

2 I collect a database of these clicks

3 I cluster what I clicked on.

4 I select from the prototypical clusters what, in my opinion, is the "averaged image" of the pattern I am looking for.

5 drill this image and go through the graph and compare the current data with the image for similarity

YES

So how many patterns did you find? Aren't the constraints for finding a similar object very tight, or did you combine similar clusters?

 
Aleksey Vyazmikin:

So how many images did you find? Aren't the constraints for finding a similar object very tight, or did you combine similar clusters?

As many clusters as you order, as many images.

You can cluster, you can't, I'm just experimenting...

And stop telling me what to do!

 
Rorschach:

Messages are broken, the file is attached.

sent, it says through the admin invitation should pass

 
Maxim Dmitrievsky:

sent, it says the invitation should go through the admin

thanks

If you remember the answers to the test, I can give you the transcript
 
Rorschach:

thanks

If you remember the answers to the test, I can give you the transcript.

The rest were easy, there were 60 pieces... the results are not available... just passed and passed

some bullshit.

 
Maxim Dmitrievsky:

The rest were easy, there were 60 pieces... the results are not available... just passed and passed

some bullshit

Below are the answers

 
Rorschach:

Below are the answers

the test is simple, apparently outdated.

1...204420452046204720482049205020512052205320542055205620572058...3590
New comment