Machine learning in trading: theory, models, practice and algo-trading - page 984
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
Maxim doesn't want to admit that he has NO model at all, which follows from the graph: learning has nothing to do with testing. His model is not retrained - it has not learned anything at all.
The extremely tempting idea of replacing the target vector with some kind of function that dynamically generates the target until it finds confirmation.
it's a working thing which is no worse than anything else shown here
And what has not been shown there is nothing to talk about
no one has a model here )
it's working stuff which is no worse than anything else shown here
I stand corrected:
it's not working crap that's no better than anything else shown here
I fixed it:
it's a dysfunctional thing that's no better than anything else that's been shown here.
good looking)
Dear R lovers, a question to you, what library in R can build a classification tree with the following features:
1. Semi-automatic construction - we decide ourselves how part of the tree will look like, what features will go on the branches and in what sequence up to a certain point, and then automatic construction.
2. Possibility to build a tree not as a binary tree - yes/no, but with branching according to a predictor, i.e. a predictor has 5 values, so we build five branches.
it's working stuff which is no worse than anything else shown here
and what has not been shown there is nothing to talk about
no one has a model here)
Please, let's eat.
I have borrowed someone else's Expert Advisor with open source code and tweaked it a bit. I copied some similar stuff in my day. If it is interesting, I can attach it. With the tester you can get very different variants.
PS.
Very interesting the ratio of profitable and loss-making trades with the profit factor 1.36
Very interesting is the ratio of profitable to losing trades with a profit factor of 1.36
In fact, this is a very good method. We try to enter, and if the result is negative, we quickly close. But one positive trade more than compensates for all previous losses.
I have shown in some thread a test of this system with only 30-40% of successful trades and quite good profit. But 90% of losing trades is great)).
Here you go, it's served.
I took someone else's Expert Advisor with open source code, did a little work. I used to rivet such things in my time. If it is interesting, I can attach it. With the tester you can get very different variants.
PS.
Very interesting the ratio of profitable and loss-making trades at the profit factor 1.36
It's a martyr from the code base on unreal ticks. Is the stooploss 5 or 4? 50 is what you have. If the 5-digit one, then you are on your own...
And what does this have to do with the Ministry of Defense, may I ask?In fact, it's not a bad method. We try to enter, and if the result is negative, we quickly close. But one positive trade more than compensates for all the losses of the previous ones.
I have shown in some thread a test of this system with only 30-40% of successful trades and quite good profit. But 90% of losing trades are great).
Yes, Sanych's loss ratio is much better than yours...
And what does this have to do with the Ministry of Defense, let me ask you?
SanSanych is tired of felling.)
Well, it's a codebase martych on unreal ticks. Is the stoploss a 5-digit one or a 4-digit one? 50 is what you have. If it's 5-digit, then you're on your...
And what does the MO have to do with it, let me ask you?Just sick of your demands to post the result!
What did you show? Some curves of unknown origin with a fig in the pocket in the form of a period of training at the end of the schedule? Not even a tester!
And here is the result, 360% in 14 months, with a decent balance chart, with a bunch of characteristics, with the original text.
And to MO has a direct result.
This is the finalized Expert Advisor from here
The input is a uniformly distributed random variable from which we get buy/sell. You see - random.
But in all exercises on this branch the error is less than 50%, and the ratio of profitable to unprofitable in the base case is 47/53.
Then why do you need the MO? And why do you need the MO in your version? After all, you don't provide any evidence that the model is not retrained! And if you are not dealing with the problem of retraining, here's a free Expert Advisor byPetr Voytenko for everyone who wants it.