Machine learning in trading: theory, models, practice and algo-trading - page 332
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
Yes, it is on the demo, it is... It's a primitive bot, which does not take into account a lot of things, but it is already able to earn... it's just a preparation for further, so to speak, research. From experience - normal systems are not written in two days, as in this case :)
Okay. Still I will wait for the results of your demo test.
Why do you need someone else's results?
IfMaxim Dmitrievsky has results, it will be the result of his KNOWLEDGE, but in no way proof of the success of the model in other hands.
MO is a tool. And either you, personally, know how to use it, or not.
The beauty of the current state of MO is that a large number of tools have become available that can be used, at least at first stage, as black boxes. Start with rattle. Six models, you can try in an hour tops. But, most importantly, you will see the full cycle of machine learning: datamining, model fitting, evaluation of results. Meanwhile, the most insignificant thing is the model itself.
Why do you need someone else's results?
IfMaxim Dmitrievsky has results, it will be the result of his KNOWLEDGE, but in no way proof of the success of the model in other hands.
MO is a tool. And either you, personally, know how to use it, or not.
The beauty of the current state of MO is that a large number of tools have become available that can be used, at least at first stage, as black boxes. Start with rattle. Six models, you can try in an hour tops. But, most importantly, you will see the full cycle of machine learning: datamining, model fitting, evaluation of results. The most unimportant thing is the model itself.
I don't need someone else's results. I need to understand - is it worth the cost?
I have a new trained model in my monitor, that I have set yesterday. I will not re-train it yet.) I lost 2 trades so far, but that's ok... if I have enough deposit I will make it to +, in the tester I had enough :) https://www.mql5.com/ru/signals/297732 traded on m15
I had a simpler model before
I tried new model in my monitor, I have a new trained model that I built yesterday and I`m not going to retrain it yet.) I lost 2 trades so far, but that's ok... if I have enough deposit I will make it to +, in the tester I had enough :) https://www.mql5.com/ru/signals/297732 traded on m15
I traded with a simpler model before
Yeah. I looked it up yesterday. The lot also surprised me.
I used to trade at m15 but I'm still looking at it. I am interested in the final result.
Thank you!
Yeah. I looked it up yesterday. Lot surprised me, too.
But I'm still checking it out. Interested in the end result.
Thank you!
I also trained on DAX and GBPUSD index, the results are also good, I can create multicurrency system with no correlation and run it directly in optimizer on the cart
And, in theory, it should work best on indexes) the regression model is perfectly adjusted to any volatility
I also trained on the DAX and GBPUSD indices - the results are also good, I can create a multicurrency system with no correlation and run it by the optimizer by the basket at once
The regression model perfectly fits any volatility, and, in theory, it should work best on indexes.
I missed something.
Have you already changed to a regression model? Which one?
I missed something.
Have you switched to a regression model yet? Which one?
It was originally in the plans... I don't know what model RNN creates internally, that's the funny thing, but it's based on regressions and probabilities :)
It was originally in the plans... I don't know what kind of model RNN creates inside itself, that's the funny thing, but it is based on regressions and probabilities :)
I read the description of RNN, which you sent me. I did not look at the code. Judging by the description, it is not a regression.
RNN is probability, regression on input :) I sent you an example with regression slope on 1 input, if you feed it to several inputs with different periods, guess what happens. :) Any market model that analyzes cyclicality