Volumes, volatility and Hearst index - page 27

 
Yurixx:

There must be a formal definition of self-similarity in this theory. I'd like to see it.

Basic:

{X(a*t), t belongs to R}=={(a^H)*X(t), t belongs to R}

{} - distribution

a>0

== equality of finite-dimensional distributions

X() is a process with stationary increments

something like this.

 
Farnsworth:

In principle, you can, but a little later, when the tedious specifics come in.

Well, it's early. I was actually interested in the methodology, but there is no urgency
Yurixx:

What, already migrated?


no, it's too early :)
 
We won't be able to think for 3 in private :o)
 

By the way, not in terms of bragging, but to start a conversation and just take TV seriously. A little bit of secrecy - simulation of trading with my system based on self-organising processes with a random structure on EURUSD, M15, CLOSE, SPRED=10 in old pips (0.0001, I got used to them in my analysis). Testing on an area of 15000 counts (150 trading days). The picture shows the charts:

  • grey thin line - the state of the deposit in pips
  • bold stepped - fixed deposit in pips as of the last trade closing ( number of trades can be seen visually)


No optimization parameters - the system is adaptive, but there are still some questions:

  • the system is very complex
  • large drawdowns
  • sometimes the adaptive subsystems go crazy (I don't have enough knowledge, I'm not a mathematician)
  • low points taken within a relatively long time of being in the market
  • no full, provable confidence that the mechanism will always adjust (hopefully it will)

FA is very interesting to me in that sense, I think it will work out to improve the efficiency of the system considerably.

 
Farnsworth:

Basic:

{X(a*t), t belongs to R}=={(a^H)*X(t), t belongs to R}

{} - distribution

a>0

== equality of finite-dimensional distributions

X() is a process with stationary increments

something like that.

Oh, I like that. And yet it means that the authors are reducing self-similarity to scalability, which doesn't seem quite correct to me. IMHO

It's kind of like a similarity of triangles. But the self-similarity not necessarily must include proportionality of a quantitative measure.

Farnsworth:
We won't be able to think for 3 in private :o)

Is that a question or a suggestion ? :-)

 
Yurixx:

Oh, I like that. Still, it means that the authors are reducing self-similarity to scalability, which doesn't seem quite right to me. IMHO

It's kind of like the similarity of triangles. But self-similarity not necessarily. should include the proportionality of the quantitative measure.

Of course not. Yuri, that is the very first definition. Are you suggesting I rewrite the books here? :о) (The list and what was in the e.v. I posted)

Is that a question or a suggestion ? :-)

Both :o) But while the production part is taking shape, I don't think it's necessary. (By the way, for the next 2 weeks - I'm on a business trip :o(

 
Farnsworth:

Of course not. Yuri, that's the very first definition. Are you suggesting I rewrite the books here? :о) (the list and what was in the e.v. I posted)

No, Sergey, don't worry so much. I was just reacting to what you wrote. It doesn't bind you to anything and my reaction is quite positive.
 
Gentlemen, I don't understand where the belief in self-similarity comes from? What is it based on?
 
Yurixx:
No, Sergei, don't worry so much. I was just reacting to what you wrote. It doesn't bind you to anything and my reaction is quite positive.
It was kind of a joke :o)
 
HideYourRichess:
Gentlemen, I don't understand where the belief in self-similarity comes from? What is it based on?
Perhaps an illusion, or maybe something to be discovered ...
Reason: