neural network and inputs - page 28

 
Vizard:
)))) everything has been badly allocated by regulators for a long time now...

what and by whom has been allocated?
 
FAGOTT:

For example, the gambling industry is SB, although we are talking about money.

There is no HOSSPs, there are HOSSPs.

Now that's interesting, this cameraman never answered that question:

1. how in the forex price flow to distinguish a group of traders exploiting a particular trading method? And we must consider that the price is formed by real traders, not by clients of brokerage companies. I hardly believe that real traders use, for example, wave theory...............

2. Suppose there is a trader/group of traders exploiting a certain trading method. How can we determine the point of entry and exit of such a group in quotes flow? After all, each theory has a number of parameters, the choice of which depends on the trader's will? For example, the elioticists may trade on different TFs, different markup and so on.

3. "Do not forget to put it all on a silver platter.)

// Oleg, why are you picking on me? Are you on your period? :)

He cited an example as a metaphor that demonstrates some logic (promising in his opinion). you start demanding a proof of this metaphor. your logic in this case looks like this: "prove this tale to me so I can take your hint."

Sounds schizophrenic to me. :)

 
MetaDriver:

3. "And don't forget to put it all on a silver platter..." :)

// Oleg, why are you picking on me? Are you on your period? :)

He cited an example as a metaphor that demonstrates some logic (promising in his opinion). you start demanding a proof of this metaphor. your logic in this case looks like this: "prove this tale to me so I can take your hint."

Sounds schizophrenic to me. :)

I'm not picking on you, I'm asking you.
If it's logical, then fine.
 
FAGOTT:
I'm not picking on you, I'm asking.
If it's logic, fine.

OK. Sorry. Good luck to you. :)

what's your own progress and approach? // I mean on the subject.

 
MetaDriver:

OK. Sorry. Good luck to you. :)

what's your own progress and approach? // I mean on the subject.


Not at all on the NS. Same nonlinear regression, but you don't have to bother choosing the type of dependence.

Linear regression for intermarket analysis gives good results, but you have to take commodities markets, precious metals, raw materials, funds. On currencies all dependencies change too quickly - the floating window should be smaller in size and the effect is canceled out.

Non-linear regression or NC gives higher prediction accuracy than linear, but the difference is so insignificant that the time and brain activity costs do not pay off

 
FAGOTT:


Non-linear regression whether NS gives a higher prediction accuracy than linear regression, but the difference is so negligible that the investment of time and brainpower does not pay off


That's probably because your brain activity is purely linear ;)))
 
avtomat:

That's probably because your brain activity manifests itself in a purely linear mode ;)))


- What are we complaining about?
- Complaining about your head.
- That's good. Lungs are breathing, heart's pounding.
- And the head?
- And the head is a dark subject, not to be examined. (С)
 
FAGOTT:

on NS - none. Same nonlinear regression, just don't bother with the choice of dependence type.

Linear regression on inter-market analysis gives good results, but you need to take commodity markets, precious metals, raw materials, fund. On currencies all dependencies change too quickly - the floating window should be smaller in size and the effect is canceled out.

Non-linear regression or NC gives higher prediction accuracy than linear, but the difference is so small that it does not pay off with time and brainwork.


you know what.... I've been thinking, and I've come to the conclusion that your conclusions can only be reasonably logical if you take only the raw quotient as an input.

But look:

the input is, well, almost standard MACD (minor tweaking, exclusively method-fitting), then one neuron with a linear output (well, normalization there doesn't count - quite linear shit), and the kernel row is dumbly dumped into an indie tester.

All this stuff is in the bottom sub-window, in the middle sub-window - obtained equity, at the top - original quotient. at the bottom - statistics.

Tweak: all this is calculated without spread (spread is deleveraged), however spread expenses can be compensated by other methods (not relevant here), in total we have 1.5 thousand normalized pips (five-digit) per day (albeit slightly less, the spread cannot be completely compensated) with a very stable chaconarity (equity curve is almost straight).

the conclusion is: you are searching not exactly with the right tools. imha.

Good luck.

 
MetaDriver:
hence the conclusion: not quite there looking with the wrong tools. imha.

success.

let's see

The real fucker will show.

 
FAGOTT:

For example, the gambling industry is SB, although we are talking about money.

There is no HOSSPs, there are HOSSPs.

Now that's interesting, this cameraman never answered that question:

1. how in the forex price flow to distinguish a group of traders exploiting a particular trading method? And we must consider that the price is formed by real traders, not by clients of brokerage companies. I hardly believe that real traders use, for example, wave theory...............

2. Suppose there is a trader/group of traders exploiting a certain trading method. How can we determine the point of entry and exit of such a group in quotes flow? After all, each theory has a number of parameters, the choice of which depends on the trader's will? For example, elioticists may trade on different TFs, different markups and so on.

Do you... You really do not understand how to distinguish a group that exploits some idea?

Reason: