Representation of an object in programming. - page 7

 

suddenly - OO exists and works well in programming languages whose syntax cannot be expressed in BNF.

 
Aleksey Nikolayev #:

Integer and comma separated list of integers (list can be empty)

<digit> ::= "0" | "1" | "2" | "3" | "4" | "5" | "6" | "7" | "8" | "9"

<integer> ::= <digit>|<integer><digit>

<list> ::= <""> | <integer><","> <list>

What exactly you want to write in this format?

Generally, such a notation |"1"|"2"|"3"|"4"|"5"|"6"|"7"|"8"|"9" would be useful to construct an example of a number of parameters, although they could also be written in comma.

 
transcendreamer #:

Um... probably then wave sequences with probabilistic rules like: if there was such-and-such an observed sequence of terms/signatures WXYZ, then further maybe ABC, ABC, BAC, ABC, CAB, CBA with some probabilities whose vector will point to the Grail in probabilistic space. 😁

Sort of)

transcendreamer #:

But it's a fundamentally Markovian model at all, which knowingly doesn't account for dependence on market background?

Any random process can be made Markovian by complicating the state space enough. In our trader's case it means the usual addition of indicator values (calculated from history, of course), fundamental data, moon phase, trader's mood and other useful information that together represent a point in the state space of the system at the moment.

 
Реter Konow #:

What exactly do you propose to write in this format?

Generally, such a notation | "1" | "2" | "3" | "4" | "5" | "6" | "7" | "8" | "9" may be useful for constructing an example of some parameters, although they can also be written with a comma.

Description of objects. This is standard practice.

 
I'm sorry, excuse me, but what did you all consume on New Year's Eve?
 
Aleksey Nikolayev #:

Description of objects. This is standard practice.

I wonder. What methods of describing hierarchies do you think would be appropriate? I mean symbolic format.

 
Maxim Kuznetsov #:

all of a sudden - OO exists and works in programming languages whose syntax cannot be expressed in BNF.

What type do they belong to in Chomsky's hierarchy?

 
Dmitry Fedoseev #:
Excuse me, excuse me, but what have you all consumed here on New Year's Eve?

I just mixed up the pills as usual, and I ran out of haloperidol a long time ago.


Aleksey Nikolayev #:

Any random process can be made Markovian by complicating the state space enough. In our trader's case, it means the usual addition of indicator values (calculated from history, of course), fundamental data, moon phase, trader's mood and other useful information that together represent a point in the state space of the system at the moment.

But it will be a fitting to the history including the false correlations, won't it?

 
Реter Konow #:

Interesting. What methods of describing hierarchies do you think would be appropriate? I mean the symbolic format.

Personally, I'm more comfortable with Haskell. Its definition of a list of items of type a would look like this

data List a = Nil

| Cons a (List a)

 
transcendreamer #:

But that would be a fitting of history including false correlations, wouldn't it?

The possibility of reducing any process to a Markovian one is a kind of mathematical fact. Another thing is that we do not have complete knowledge of the process, but only one realization of it. Therefore, there is no guarantee that the state space we choose in Markovization will be correct. In practice, this can be seen, for example, in attempts to apply HMM.