Machine learning in trading: theory, models, practice and algo-trading - page 2341
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
they are kind of obvious. The sad thing is that it is not clear what is needed to understand what the drain states differ from the normal state. I would work around them. In terms of a filter. )
The bias in the model starts, the cause-effect goes to the f...u., the market changes. Predictors in the form of increments are not able to drag for a long time.
you can even see for each individual predictor what happenedThe model starts to change, the cause-and-effect model goes to hell, and the market changes. Predictors in the form of increments are incapable of dragging for a long time.
you can even see for each individual predictor what happenedLike a pattern to find in increments on plummets. or in changes of predictors).
like a pattern to find in the increments on the plums.
If you combine it with a pattern not on plums, then on average it will be random and the model will not learn properly
If you combine it with a pattern not on plums, the average will be a random and the model will not learn properly
And why connect, the goal is to isolate and understand whether these are the same states or not. If they are the same, then we can isolate something. If states are random, or there are many states and they can't be grouped.
And why connect, the goal is to isolate and understand whether these are the same states or not. If they are the same, then it is possible to distinguish something. But if the states are random, or there are many states and they cannot be grouped.
I don't see anyone trying to do something like this
I don't see anyone trying to do anything like that.
Cutting a row... to isolate something... I haven't seen it either. Although in spectral analysis they sometimes do this.
If you combine it with a pattern that is not based on losses, then on average it will be random and the model will not learn as well as it should
then just cut the losses - if the TS has lost more than the specified value of the loss - close positions and do not trade for 24 hours
in 99% of cases you get a more or less stable TS, if after such manipulations TS doesn't pass the test, then the original TS has outlived the losses
then just cut the losses - if the TS has fallen below the loss set value - close positions and do not trade for 24 hours
in 99% of cases you get more or less stable TS, if after such manipulations TS does not pass the test, then the original TS has outlived the losses
on that data no matter how you slice it - the result will be the same )
Cutting a row... to isolate anything... I haven't seen that either. Although in spectral analysis they sometimes do that.
Mix in different pieces of history, I don't know how anymore.
Mix in different pieces of history, I don't know how anymore.