Machine learning in trading: theory, models, practice and algo-trading - page 2123
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
Stupid about what?
You're writing something crazy.
it seems like we already defeated those who yelled that the forward should only be on the front, and a long time ago
just like someone yelling, sputtering snot, that convolutional networks are not used for time series (about a year ago, maybe more)
You're writing something crazy.
OK, don't listen to a fool)
I think we've already defeated those who yelled that the forward should only be up front, and a long time ago.
How have you won? you just talked and decided)))) a serious approach
First explain yourself, and then me, how is it possible to predict the past point from the current point,and we know it, but the future onedoes not work ...
Although if you look from your point of view there is no difference what point to predict the past or future ... because the model hasn't seen either one.
just like someone yelling, sputtering that convolutional nets aren't used for time series (about a year ago, maybe more)
what a schizophrenic))
How did you win? We just talked and decided)))) A serious approach.
Explain first of all to yourself and then to me how it turns out that from the current point the past point can bewell predictedand we know it, but the future does not work ...
Although if you look from your point of view there is no difference what point to predict the past or future ... the model hasn't seen either one
you're a schizophrenic)
it's already been explained, check out the forum
It has already been explained, browse the forum
Explain it to me, I'd appreciate it...
You know that this swamp can be leafed through until old age, and not the fact that it is there
You can explain, sorry, I'd appreciate it...
You know you can flip through this swamp until you're old, and it's not a fact that it's there.
fiches do not look that deep into the past, at most a few days ago. The treine is half a year, the test is 20 years old.
the treyn at the end, so that the model is as close to the current moment as possible
you're suggesting we train for the year 2000 and trade today.
all, goodbye.
fiches do not look that deep into the past, at most a few days ago. Train half a year, test 20 years
the trainee at the end so that the model is as close to the current moment as possible.
You're suggesting we train for the year 2000 and trade today.
Okay, goodbye.
logic is lame, but ok
If the test is 20 years old, then there's no reason to make it closer to the current moment, and if you want to make it closer to the current moment, then there's no reason to test it for 20 years )) you may find the best models in a shorter period of time
logic is lame, but ok
if you've been testing for 20 years, there's no need to get as close to the current moment as possible, and if you want to get as close to the current moment as possible, there's no need to test for 20 years )) you can find better models in a shorter time
why do you ask if you keep bending your bullshit?
Why did you ask if you're still talking nonsense?
I don't know, I'm in a fucked up mood.
How Ivakhnenko recommends to divide, so that the model is trained properly
http://www.ievbras.ru/ecostat/Kiril/Library/Book1/Content393/Content393.htm