Whether there is a process whose analysis of one part does not allow predicting the next part. - page 5
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
I remember the first time I was flattened by the existence of these fast substitutions, when at the institute I was explaining hardware FFT calculators, there just a bunch of such tricks are used. There's also such a thing as "iterative methods" - but it's hard to understand how, for example, trigonometric expressions can be calculated in just a few simple operations...
О!. By the way, advise the second Alexey (Mathemat) on some iterative fast convergent algorithm for calculating Pi. He already has one, but it's very polynomial...
// On the other hand - he probably doesn't need an efficient algorithm, it is a test case for performance... :)))
Of course you don't. I know the fast convergence one myself. I need billions of operations - to count and count to the end.
As if I didn't know that the sum of inverse squares is slow to add up...
Hello.
I suggest that the esteemed community come up with a process that cannot be predicted (so that no money can be made on this prediction). At the same time, the process should not have stationary stat-characteristics over time.
Not stationary. By definition.
Let's get this straight. We can fight non-stationarity. We get a model with variable parameters. Deterministic.
Clarify. We obtain a model where parameters are stochastic.
Clarify. Obtain a model where the functional form changes (was linear and became quadratic or vice versa, etc.).
Bottom line: the process has a breakpoint. The publications of the last 3 years are exactly on this topic. The problem is obvious. The shift can be identified on history. But what if the shift started at the last bar?
We toss a coin, heads tick up, tails tick down... Instead of a coin, MathRand()%2.
When checking for parity, Rand algorithms quickly degenerate and start oscillating, so this tick generation method should not be used. The vulnerability is found in many C-like compilers, including MQL4.
I remember the first time I was flattened by the existence of these fast substitutions, when the institute explained hardware FFT calculators, where just a bunch of these features are used. And then there's such a thing as "iterative methods" - it's hard to grasp how, for example, trigonometric expressions can be calculated in just a few simple operations...
Assembler rules!
When checking for evenness, Rand algorithms quickly degenerate and start oscillating, so this tick generation method should not be used. The vulnerability was found in many C-like compilers including MQL4.
Well, here's the rubbish...
And here is my chart, bars at 10,000 ticks, closing prices are shown:
MathRand() function from mql4 is used. Guess how it is used.
There are indeed repeating waves with a division gap.
There are a lot of options really.
There's more.
int d = MathRand()%7 - 3; c+=d;but that's beside the point.
The question is, can you make money on this row? What about that one over there? And on the third from the left side...?
That is, to make a profitable MTS "earning" (having a stat advantage) on this row. Playing with the spread, self-sabotage, albeit minimal.
I would have liked such a forum game. Kinda like quacking "random" (pseudo, right?) processes, with the help of robots. :)
And then make new "random" rows (with the posted code of row generator. immediately posted or later - the question of agreement) and shake them again.
Fortunately, tester in four allows to import quotes from particularly gifted sources...
;)
I'd love a forum game like this. Sort of quack "random" (pseudo, right?) processes, with the help of robots. :)
;)
I'm in the process of developing similar identification methods. This is a non-trivial task. Standard statistical methods are not suitable here. Perhaps in some time I will post something in the "Hurst Index" thread.
The question is - can this row be "earned"? And on that one over there? And on the third one on the left side...?
I.e. to make a profitable MTS "earning" (having a stat advantage) on this row. Playing with the spread, self-sabotage, albeit minimal.
Unfortunately any forecast can only rely on a deterministic component. If there is no deterministic component, any prediction, and therefore any earning, becomes impossible.