Machine learning in trading: theory, models, practice and algo-trading - page 3026
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
any names
If the names of columns of the table of the general sample are taken from the tree, it's fine.
You can think about speed later, if it will work at least efficiently.
If the column names of the general sample table are taken from the tree, it's fine.
You can think about speed later, if it works at least efficiently.
You've been picking leaves for 500,000 years. Have you found any good ones? At least 10 of them.)
I published the results. Yes, there are normal variants. But I haven't done it for three years.
Another thing, my experiment with leaf selection is limited to sampling alone.
I published the result. Yeah, there are some good options. But I haven't done it in three years.
Another thing, my experiment with leaf selection is limited to sampling alone.
What do you mean I haven't done it for 3 years? What do you do?
.
Globally - quantum tables and was engaged in. A lot of tests and experiments have been carried out, including on different samples.
I was engaged in quantum tables globally. Many tests and experiments have been carried out, including on different samples.
and in what form do you shove the rules into catbust? or does it not participate there at all?
What kind of rules do you put in the catbusters? Or does he not participate at all?
In binary form. The column is the number of the rule, and the value is "1" - the rule worked and "0" - the rule did not work. And the target is just like the main sample.
That's one way to aggregate everything. But, CatBoost doesn't do a very good job here, it seems to me - very rarefied data.
In binary form. The column is the rule number, and the value is "1" - the rule worked and "0" - the rule did not work. And the target is the same as in the main sample.
That's one way to aggregate everything. But, CatBoost doesn't do a very good job here, it seems to me - very rarefied data.
Also, the rules are one way buy/sell. Are the stops just being adjusted to fit them? If you don't put them in the bousting.
I think just to generate at once a check bot and check the necessary rules through the tester/optimiser.and the rules are one way buy/sell. Do the stops just match them? If you don't shove them into the bousting
In the old approach, the results of which I was demonstrating, there were 3 class labels - "1" - buy, "-1" - sell and "0" - do not trade.
Now I use two labels "1" - trade and "0" - do not trade the column name is "Target_100". The direction is defined by a separate column "Target_P", and for the financial result of buying and selling the corresponding columns "Target_100_Buy" and "Target_100_Sell". Also in the sample there is an auxiliary column with date "Time".
In general, the tail of the sample contains all these columns and it looks like this