Machine learning in trading: theory, models, practice and algo-trading - page 3535

 
Aleksey Vyazmikin #:

I don't have a queue of investors to rely on other people's money, and I'm more responsible for other people's money than my own.

The investor takes the risks, you take the risks of sitting around and dumbing down in programming. Money is a tool like everything else.

 
Maxim Dmitrievsky #:

The investor takes the risks, you take the risks of sitting around dumbing down programming. Money is a tool like everything else.

Sounds good.

I would also give up programming ideally - just ideas and algorithms.

 
Aleksey Vyazmikin #:

Sounds good.

I would also give up programming ideally - just ideas and algorithms.

Well, that makes sense. Investor to investor, developer to developer.
I wouldn't do anything at all
 

A bizarre picture on the graph is obtained - y is the number of selected quantum segments for all iterations on the predictor, and x is the percentage of stable ones.

I don't understand how to interpret the result.

 

resembles moire in modular operations...or in short - properties of integers, divisibility/ multiplicity and so on. (accumulated double errors also look like this).

 
Maxim Kuznetsov #:

resembles moire in modular operations...or in short - properties of integers, divisibility/ multiplicity and so on. (accumulated double errors also look like this).

Yes, there's an obvious bug somewhere.

Back in the thousandth year I told you that you should look at the distribution of weights of the network, but the local audience is all giggles and giggles.

Just by looking at the distribution of weights you can already judge a lot of things, including the fact that this is not a working system (or working).

In general, the distribution of weights is one of the litmus papers, you can judge by them even without doing OOS.

 

To a bum, bum's bum.

A freeloader - cholyavego.

A pisser is a pisser.

Some kind of matrix? No, just nonsense, the distribution of desires and opportunities, yeah, the same omnipresent distributions.

 
To the optimiser, optimizdnuty.
 
To a dickhead, dickheads are dickheads.
 
Aleksey Vyazmikin #:

A bizarre pattern on the graph turned out - y is the number of sampled quantum segments for all iterations on the predictor, and x is the percentage of stable ones.

I don't understand how to interpret the result.