From theory to practice - page 742

 
Novaja:

This is how my system earns SB, who has any say?


Equity work on Sat is the same Sat.
 
Yuriy Asaulenko:

There was a topic recently, I don't remember whose it was. It was suggested to take an array of trades of a successful TS or signal from a tester or real, and use them to train the TS. By the way, the idea makes perfect sense.

Where is "going through all the indicators and their settings"?

The fact that the system works on the real market, and not on history - that already says something.
 
Олег avtomat:

This seems strange only because it contradicts established stereotypes. But these stereotypes have been manipulated into the minds of the masses. There is no evidence (of impossibility) anywhere.

And there is no evidence anywhere that it is possible
 
multiplicator:

you need to SLAVE(0;1). If 0, then -1. If 1, then 1.

Yeah, I didn't figure out that only integers are output there, until I wrote it in Excel. It's more reliable, though, to deal with float doubles, and make a coin out of them.

 
multiplicator:
the equity of working on the sb is the same sb.
why would it be? :O
 
_o0O:
why? :O

With the fact that patterns cannot just appear out of thin air as a result of magical manipulation. They either exist in the original process (and then they can be found), or they do not exist (it is impossible to find something that does not exist).

SB does not contain regularities by definition, hence any derivative of SB does not contain them either.

 
secret:

With the fact that patterns cannot just appear out of thin air as a result of magical manipulation. They either exist in the original process (and then they can be found), or they do not exist (it is impossible to find something that does not exist).

SB does not contain regularities by definition, hence any derivative of SB does not contain them either.

An extremely simple example: we apply MM by martin, are you claiming that equity will turn out to be the same SB with the same characteristics as the original SB? There will be a NEW pattern that didn't exist, won't there?

Again, why would it? (this is a very tolerant analogue of the word s*u*ya*i).

I can, and you can, if you think about it, beat the original SB in a way that Mama Bernoulli would never know.

 
secret:

SB does not contain patterns by definition, hence any derivative of SB does not contain them either.

This is wrong. Draw the MAshka - it is derived from SB, and you will get a lot of regularities. You may build a tester's grail on them at once).

 
Yuriy Asaulenko:

There was a topic recently, I don't remember whose it was. It was suggested to take an array of trades of a successful TS or signal from a tester or real, and use them to train the TS. By the way, the idea makes perfect sense.

A very sensible idea. But we must take an array of deals of a (very) successful trader who trades by himself, without robots and not following a rigid strategy. The market is simply analyzed using available tools and makes decisions.

But this input is hard to find, the probability is about 0 :-) And how to teach TS on it is also unclear.

 
Maxim Kuznetsov:

But you can't find such an input, the probability is about 0 :-) And it's not at all clear how to train the TS on it either.


With or without a robot - it doesn't matter.

How to teach it is clear, but I'm not saying that it's easy. But you will never find it, that's for sure).

Reason: