Machine learning in trading: theory, models, practice and algo-trading - page 1739

 
Maxim Dmitrievsky:

No, of course not.

Well then go for it!!!

Truth I do not quite understand why it works, I tried similar, took the target and mixed it in the cluster, but nothing good came out of it

 
Maxim Dmitrievsky:

What the hell, you should either understand the algorithms, or just leave everything in python. I'd like to use an mt5 tester for further manipulation

another example, train and test for the year


i don't understand what you are doing.)

What are you combining on what TF?
 
Maxim Dmitrievsky:

So do I.

show me the price chart

and the path below will be a cluster histogram

as an indicator

Or paint the chart itself with the colors of the clusters, it would be even better

 
mytarmailS:

The pictures are great, but I don't understand what you're doing.)

What are you merging there on what TF

a couple of different increments, then glue the returns on each predicted cluster like you did when gluing

 
Maxim Dmitrievsky:

put ina couple of different increments, then glue returns on each predicted cluster, like you did when gluing

I'll try it now.

What's a n ?




I've split it 10 ways...

But this isn't a prediction, it's just a statement of fact.

 
mytarmailS:

Let me try.

What's a pl?

I broke it down into ten...

But understand, it's not a prediction, it's just kind of a statement of fact.

test keys.

and I'm asleep.

 

This is bullshit, maybe I did something stupid in the code.

 
mytarmailS:

This is bullshit, maybe I did something stupid in the code.

Make three clusters. Where the hell did you break the cloud into 10 parts, there will be the hell

and think about why 3 is necessary, in the context of increments... gone
 
Maxim Dmitrievsky:

Make three clusters. Where the hell did you break the cloud into 10 parts, there will be the hell

and think about why 3 is necessary, in the context of increments... gone

Here's 5, test the top one.

I will make 3 now





 
mytarmailS:

Here are 5, test tops

I'll do three.

and color the clusters on the graph later, if you can... it's hard in python

to see the length of the clusters, how they alternate, etc.

Reason: