From theory to practice - page 365
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
Gentlemen!!!!!!!!!
It's coming to a close, isn't it? Finita la comedy, as they say.
I assure you that Erlang's flows are the key.
Here, literally, I just checked this week's AUDCAD quotes.
1. No time intervals help to read the quotes evenly. All the same, on M1, M5, etc. there is no symmetric distribution, even remotely resembling normal, or Laplace. Impossible to get, do what you want.
2. When passing from simple flux to Erlang flux of order 300 (something like M5), Laplace distribution for increments is confidently observed.
I haven't checked further yet.
Regards,
Schrödinger's cat.
i.e. the exponential readout can be removed, or is it still primary and then Erlang?
I.e. can the exponential readout be removed, or is it still primary and then Erlang?
It turns out that it is possible to set an HF generator with Erlang distributionhttps://en.wikipedia.org/wiki/Erlang_distribution of order 300 and read tick quotes at these time intervals. Smaller orders can be not considered - the transition to Laplace distribution is observed only from 300.
Unfortunately, I don't know of such a "Laplace process" as opposed to a Wiener process. But, it should still make the problem much easier to solve.
It turns out that it is possible to set at once an HF generator with the Erlang distributionhttps://en.wikipedia.org/wiki/Erlang_distribution of order 300 and read at these time intervals the tick quotes. Smaller orders can be not considered - the transition to Laplace distribution is observed only from 300.
Unfortunately, I don't know of such a "Laplace process" as opposed to a Wiener process. But, it should still make the problem much easier to solve.
And there is also q-gaussian distribution, can it somehow be relevant here? There is something about entropy and about everything, it's just that the codes are already there :)
I haven't understood anything from the article yet
With Close on the minutes, everyone works. Here you are in competition with everyone, even Papuans. And in Erlang flows you are alone, and with the Laplace distribution with its known quantile function.
With Close on the minutes, everyone works. Here you are in competition with everyone, even Papuans. And in Erlang flows - you are alone, and with the Laplace distribution with its known quantile function.
(Mm-hmm. If you refine the distribution by 2-3% - you won't even notice these errors on the graph.)) Here you have no advantage, not even over Papuans).
With Close on the minutes, everyone works. Here you are in competition with everyone, even Papuans. And in Erlang flows you are alone, and with the Laplace distribution with its known quantile function.
Laplace distribution, Exponential as a special case of Erlang distribution at k=1, Gamma distribution, analog of continuous geometric and simple Poisson flow and a special case of Weibull distribution have a key feature - lack ofmemory. The Laplace distribution, although tending towards the normal distribution, has denser tails.
While A_K2 is fiddling with Erlang flows, we've all had it here for a long time). We take minute data, say Close, and already have Erlang flow of about 90-100 order. And all distributions are where they should be. What is there to think about? We need to shake on it.
You won't get astronomical time, it will shift, it's operating time.
Laplace distribution, Exponential as a special case of the Erlang distribution at k=1, Gamma distribution, analog of the continuous geometric and simple Poisson flow and a special case of the Weibull distribution has the key property of nomemory. The Laplace distribution, while tending towards normal, has denser tails.
Tails are not memory. Memory is the dependence of the next increment on the previous one.
Distributions do not carry the slightest information about the presence/absence of memory - for that you have to look at conditional distributions or autocorrelation, which are essentially the same thing.
A simple illustration: I can shuffle any series of gradients (swap gradients randomly). The memory may or may not appear. But the distribution remains unchanged.
Citizens suffering from this problem, google and study the basics. Otherwise it is ridiculous to read you.