Machine learning in trading: theory, models, practice and algo-trading - page 3130
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
How do you know what's done by whom? That's a bit of a boast. I have signs that take into account movements over the last three months.
Taking into account movements on different TFs and seeing levels on different TFs are not the same, both in essence and complexity...
You're getting a lot of rude, what's wrong? Did you stop taking your pills or something?
Of course it's obvious - there's a target and some logic connecting them.
Earlier in a separate thread I showed that it is possible to build different models from the same data - I get about 30% profitable and 70% unprofitable. That is why I think that the method of estimation through the model is not very reliable.
Taking into account movements on different TFs and seeing levels on different TFs is not the same thing, both in essence and complexity....
You've become so often rude, did something happen, did you stop taking pills or what?
Are you worried about me? Maybe I should put in an account to top up the money for the pills.
As a matter of fact I write - some kind of a widespread feeling of their uniqueness in the participants of this branch. How can you think that others do not do so.
On the contrary, I found that some of the methods that I used are used by other people to solve similar problems. But I got the information too late.
I have the same data, adjusted for the sid
.
That's the point, the syd allows for different model building. So how can one claim that the built model is "correct" - that's my claim for these methods.
That's the point, the syd allows different ways to build a model. So how can one claim that the constructed model is "correct" - that's my claim for these methods.
And what are the other methods?
I answer, so as not to waste time. These are statistical tests, for example through randomised experiments. Sometimes useful too, this is the very basis of kozul, before moving to ML.
Can you show me something like that?
If it's that common.
I thought so.
I highly recommend you to reconsider the very approach to feature construction. Functional transformations do not work here, the market is NOT a time series!
Here is the M1 entry point
here is the entry reason and entry price H1
=========================================
And all your signs are doing is looking for patterns/curves in the last 5-10 candles.
It's more complicated than that.
Finally!
So, we have a start.
What's next?
And what other methods are there?
To answer, so as not to waste time. These are statistical tests, for example through randomised experiments. Sometimes useful too, it's the very basics of kozul, before switching to ML.
You're partly right, except I don't understand the philosophy of shuffling everything around - it certainly works if there is just no irrecoverable drift, for example with cycling.
To begin with, I would like to classify different types of drift, and then work with them individually - if the cause is known, then we can think of a way to eliminate it. And if not eliminating, then detecting (detection).
You're partly right, but I don't understand the philosophy of shuffling everything around - it certainly works if there is no irretrievable drift, for example with cyclicality.
To begin with, I would like to classify different types of drift, and then work with them individually - if the cause is known, then we can think of a way to eliminate it. And if not eliminating, then detecting (detection).
randomisation removes the bias between the test and control, after that the predictor impact is evaluated
If you don't remove the bias before that, it will be an associative relationship, not a causal one.