Machine learning in trading: theory, models, practice and algo-trading - page 2144

 
Valeriy Yastremskiy:

I want to ask you - have you somehow managed to achieve a guaranteed synchronization of news time with quotes (on an arbitrary interval of history)?

 
Maxim Dmitrievsky:

I guessed it quickly) but no one else...

Cool stuff!!! Strong :

In MNIST, for example, we achieve 99.14% accuracy with only 10 labeled examples per class with a fully connected neural network- a result very close to the best known results with fully controlled approaches using all 60000 labeled examples.

But I would still recommend double-checking for errors, just in case...

 
mytarmailS:

Cool stuff!!!

But I would also recommend double-checking for errors, just in case...

I don't make mistakes, only conceptual ones

don't shout much, just study it quietly )
 
Maxim Dmitrievsky:

I don't make mistakes, only conceptual ones

Yeah))) how many that bot put and removed? ))

 
mytarmailS:

Yeah))) how many times have you installed and removed that bot? ))

there was a completely different approach, and here already worked out all the f-i

 
Maxim Dmitrievsky:

It was a completely different approach there, and here all the f-i's have already been worked out.

well, then go full steam ahead!!! put the bot on the demo

 
mytarmailS:

Well then full speed ahead!!! Put the bot in the demo.

Mon, it's Friday already.

 
Maxim Dmitrievsky:

Mon, it's Friday already

Well, not much time yet

 
mytarmailS:

Well, there's still plenty of time.

And when to drink wine?

 
Maxim Dmitrievsky:

And when to drink wine?

In the evening...

That's what I do, like all normal people....

I start with beer in the morning and wine in the evening on Tuesday...

Reason: