Machine learning in trading: theory, models, practice and algo-trading - page 2593

 
Maxim Dmitrievsky #:
It's not my fault you only recently discovered it.
Ahahahaha, right.
 
What's wrong with someone absorbing information from open sources, analyzing it, and using it for their own purposes?) Information and ideas are nothing until they take the form of results.
 
mytarmailS #:
Ahahaha, exactly
Sounds like hysteria ) Nobody's using your 'ideas', calm down.
 
Replikant_mih #:
What's wrong with someone absorbing information from open sources, analyzing it, and using it for their own purposes?) Information and ideas are nothing until they take the form of results.
It's a long story.
 
Maxim Dmitrievsky #:
Sounds like hysteria.) Nobody uses your 'ideas', calm down.

That's it, I'm calm))

 
Aleksey Nikolayev #:

I asked a question if it is possible to restore the OP, I wrote to me that it is possible with the help of an algorithm Metropolis Hastings, this is a kind of simulation montecarlo or something like that, in short, I do not know a bit about it, there are several packages for R...

Can you help me with that?

 
mytarmailS #:

I asked a question if it is possible to restore the OP, I wrote a man that you can use the algorithm metropolis hastings, it's some kind of simulation montecarlo or something like that, in short, I do not know a bit about it, there are several packages for R...

Is there any way you can help with that?

I don't really understand the connection between the posterior distribution and the target function.

 
Aleksey Nikolayev #:

I don't really understand the connection between the posterior distribution and the target function.

Me even more ((
 
mytarmailS #:
I'm all the more so((
You need to pull all the model weights at each iteration of the training and plot the model scoring dependencies at each iteration on the weights. You'll get an optimization hypersurface that won't give you anything. Guess why? You have three attempts. You have even more ways to fail, especially you will be hurt physically and mentally in case of bouncing, but you're used to it 🤣🤣🤣🤣 But in other cases there is an ambush waiting for you too.
 
Maxim Dmitrievsky #:
You need to pull out all model weights at each iteration of training and build the dependencies of model scoring at each iteration on weights. You'll get an optimization hypersurface that won't give you anything. Guess why? You have three attempts. You have even more ways to fail, you'll have mental and physical pain in case of bouncing, but you'll get used to it 🤣🤣🤣🤣 But in other cases there is an ambush waiting for you too.
God's criticism again.