Machine learning in trading: theory, models, practice and algo-trading - page 1124
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
And yet, what difference does it make that the forward is behind the trayn? If the samples do not overlap. They are just sets of points, not connected in any way (well, maybe connected to a small depth because of the lag of the predictors).
You can ask him :)@fxsaber or how to make an invitation hereThe @ invite didn't work. I don't understand what this is about. It feels like the posts were deleted.
The @invitation didn't work. I don't know what this is about. It feels like the posts have been deleted.
Let's just dedicate the evening to your genius :)
there was that he wrote that you can't do a forward on the back of a tray, as on your screenshot... but then we decided that it's unreasonable
It feels like the posts have been deleted.
Alas, someone in this thread has taken this way of writing posts and everyone has picked up on it... wrote, read-ster.... stupid, but that's how we all do it )))
There was about that he wrote that you can't do forward behind trayn, like on your screenshot... but then we decided that it's unreasonable.
Maybe a little off-topic for the last conversation, I'm with my own turd))
Here is the process:
Returns process
Is it possible to work with a process like this?
PS Decided to add:
Stat characteristics and distribution:
)))))) You're right about most bulls and bears alike, as well as those who only optimize the error on the lern and then also show their little... I mean backwards as an argument. I wish you the "profitable trading" you speak of. Do you have a PAMM by any chance? Or a "successful trading" course taken from a rental car dealership?
Once again, the backtest is shown as an argument that your assertion about the randomness of the EA is false, because even a snot-nosed schoolboy knows that the randomness is impossible to purposefully solve the problem, such as a profitable trading of the EA in the tester.
If you want to defend your case, then instead of empty trolling provide your evidence with examples.
And the "successful trading" trick seems to be the professional headache of algotrader
Forum on trading, automated trading systems and testing trading strategies
Machine learning in trading: theory and practice (trading and beyond)
2018.10.20 11:39
well, I personally communicate with specific people, for my own benefit or fun at the moment, but I am not going to teach potential competitors en masse, so I'm wiping behind me, who else also does not know and do not assume why.
AAHAHAHAHA))))))) Exactly what even a snot-nosed schoolboy can see that even in SB, it is ELEMENTARY to fit the training sample, so this is NOT an ARGUMENT at all, but an IDIOTISM.
There is no need to adjust for the training sample at all (optimize), normally the sample is divided into a train, validation and test. We learn on a train but optimize error on validation, as full cruteness to optimize the error in the same data, it... well, except for some academic experiments, and then all this is run on TEST to get results.
I'm shocked that... it's not even schoolboys but imbeciles, they optimize something like mashka on a sample and show the "result" on it as an argument, it's a shame.
What the hell is the point of all this? I'm not even talking about the week sample size, maybe that's why our near-market operator doesn't show the forward one... what the fuck to say))) Laugh and sin
I hope you don't mean me about the near-market guy???? because I also train on 4 weeks, but I've been trading the robot for a long time and you can't call me that impudent way. I am a practitioner.
Well you don't have a website with grails and weight loss training and you don't argue with curves on backwards, no it's not about you.
Phew... that's a relief. Today I remembered again about the promised video for the apology in graalespiharstva. When I have an indisputable trump card in my hands I think I will record it... :-)
AAHAHAHAHA))))))) Exactly what even a snot-nosed schoolboy can see that even in SB, it is ELEMENTARY to fit the training sample, so this is NOT an ARGUMENT at all, but an IDIOTISM.
There is no need to adjust for the training sample at all (optimize), normally the sample is divided into a train, validation and test. We learn on a train but optimize error on validation, as full cruteness to optimize the error in the same data, it... well, except for some academic experiments, and then all this is run on TEST to get results.
I'm shocked that... it's not even schoolboys but imbeciles, they optimize something like mashka on a sample and show the "result" on it as an argument, it's a shame.
What the hell is the point of this? I'm not even talking about the week sample size, maybe that's why our near-market operator does not show the forward, what the fuck... what can I say))) laugh and sin
You wrote a false statement again - I showed the forward test in the first place and I don't trump my posts, unlike some.
https://www.mql5.com/ru/forum/86386/page1120#comment_9071338
In addition I gave the invest account password and here is a fresh screenshot:
Strangely enough, but profits continue to grow and if you have nothing to explain it, it is better not to write at all, I withdraw the question I asked, make others laugh.