Machine learning in trading: theory, models, practice and algo-trading - page 142

 
SanSanych Fomenko:

Well, now you're offended...

My goal is to steer the conversation in a practical direction, not to offend anyone in any way...

So far we have scattered bits and pieces.

Your academic approach.... For me the value of your calculations is unquestionable, but .... I expressed my doubts above.

I follow carefully the worksof Vladimir Perervenko. I have never seen evidence that the models are not overtrained. The last link. The importance of variables is determined by an algorithm of one of the tree variants. But trees, because of the convenience of available noise values, tend to use these noise predictors more often and as a result noise pops up in the importance estimation...

So you have to start with algorithms to remove the noise predictors. All other steps without these make no practical sense, as all model estimates cannot be extrapolated into the future.

Then training the model in windows, and the width of the window must be justified somehow. Then using the trained model in a pre-selection of predictors for the working window....

Something like this.

I am not offended. I am collecting information for further use. You didn't spell out your approach, no graphs again. All that would have been interesting. And in general, a graph in MT before optimization of the robot and after on the work period. All that would have raised interest. Otherwise... Just a dry idea.
 
Alexey Burnakov:
I am not offended. I gather information for further use. You didn't describe your approach, without graphs again. All that would have been interesting. In fact, a chart in MT before and after the optimization of the robot on the working period. All that would have raised interest. As for now... Just a dry idea.

I have a working one year old robot. It uses rf. What share of the robot's performance belongs to rf I don't know. But besides rf in the robot there are many other things.

So I do not know how to fulfill your wish.

I'm writing about an idea: upgrading the robot with machine learning inserts.

 

please tell me how to remove the separator "-" from the vector with the date

head(date)
[1] "31-10-2014" "31-10-2014" "31-10-2014" "31-10-2014" "31-10-2014" "31-10-2014"

to make it work

head(date)
[1] "31102014" "31102014" "31102014" "31102014" "31102014" "31102014"

I can't find anything.

 
I will be gone for two weeks. I am flying to Turkey.

For the dates, see help. You can do everything there.

Or just convert to character and remove/replace dashes.

When I come back, I'll process the results of the experiment and post them.
 
Alexey Burnakov:
I`ve been away for two weeks. I am going to Turkey to rest.

For the dates, see help. You can do everything there.

Or just convert it to character and remove/replace the dashes.

On my return, I'll process the results of the experiment and post them.

It is "character". I do not know how to remove it ...

P.S.soft landing and have a good rest

finished ...

all rebound, got it all figured out... thanks for the tip

 

You have amazed me once again.

And this is the level of thinking of a trader?

:(

 
Vadim Shishkin:

You have amazed me once again.

And this is the level of thinking of a trader?

:(

What and who are you talking about?
 
mytarmailS:
What are you talking about and who are you talking about?
He's trolling again. And again in the bathhouse. Have fun.
 
Alexey Burnakov:
He's trolling again. And again in the bathhouse. With a little steam.
Symmetrical. :)
 

I had a discussion with Renat on a neighboring thread about the fate of the R language in MKL4/5. The solution and direction of development of MKL is now clear.

Good luck