Machine learning in trading: theory, models, practice and algo-trading - page 1567
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
Max, you just said what the Wizards don't say. Either they take the conversation to nowhere, or they drag the lop-eared sufferers into the abyss with false insinuations.
So it was, so it is, so it will be.
Just getting to the bottom of the usual random market process is extremely difficult. Market BP is the sum of 2 random processes, resulting in a very complex non-Markovian series. It is not possible to take it by the modern mathematical apparatus.
Therefore, the real seekers should have only one goal - to see these 2 subprocesses, or to reduce the initial non-Markovian series to the Markovian one.
That's it.
Goodbye, Max - I don't live here anymore. This is just my Shadow...
vergüenza ajena
Moreover, it turns out that in the market it is more difficult to make money than in the SB
Of course it is more difficult, earning on SB is 50/50, and on the market (50 - spread - commission).
Certainly it is more difficult, earnings on SB are 50/50, and on the market (50 - spread - commission).
Of course MO is a bit not my topic, just to note - spread + commission are removed unfortunately only by martingale. It's small at first, but it's brutal.
This is provided that the other is at least minimally positive mathematical expectation and permissible variance. Then it is possible to cut off the excess by martin.
While you build models and strategies, forget about spreads altogether.
If possible, I will smoke, so far from the first post of AK, I have the impression that all such statistical studies are based on the assumption that there is a history-sample, which is large enough to look for something there.
If you look at it philosophically, you can see that it is impossible to prove that the sample is not TOO small, so much so that any statistical research on it makes no sense.
Simply put - whatever regularities we have not found in the entire trading history, the probability that at the next moment (period) of time something will happen that will change the regularities in the entire history, it is not less likely than that - the regularities will not change. And it is impossible to refute it.
This is probably true of any random process. Apparently, what matters is the repeatability and "it" counts, which gives hope to those who suffer.
A simple Random walk is filled to the brim with such situations, but in the limit, of course, there are none.
It turns out that some particular implementation of SB always has stable patterns. Maybe it depends on initial conditions or who knows what else. In the market it is somewhat more complicated, there may be a superposition of several processes, as Alexander writes about. Further on, my knowledge is not sufficient to deduce anything theoretically, I will only look at it in practice.
Not a bad illustration for a random walk, in my opinion:
http://investazy.com/blog/280.php
Perhaps this is true of any random process. Apparently, what matters is the repetition and "it" counts, which gives hope to those who suffer.
A simple Random walk is filled with such situations to the brim, but, in the limit, of course, there are none.
It turns out that some particular implementation of SB always has stable patterns. Maybe it depends on the initial conditions or on who knows what else. In the market it is somewhat more complicated, there may be a superposition of several processes, as Alexander writes about. Further, my knowledge is not enough to deduce anything theoretically, I will only look at it in practice.That's right, also a nuance comes to mind - how exactly do all such researchers get SB? (maybe it was somewhere but I missed it or didn't get to it)
In the sense of how random it really is.
The biggest poker site in its time gave a description of its RTC engine (it cost a lot of money by that measure), and there they used not only high-precision sensors of temperature, pressure in several places of the server room and other similar things, but even data on last mouse movements of users at a certain poker table. I hope this gives an idea of the level of randomness of GSR, which they estimated at four nines (That's all!).
And they came to this not because they wanted to show off their budget. And because more simple models of RNG have practically extremely low resistance to hacking (pattern finding), and there were many precedents of poker rooms' RNG hacking
back in the day. This is more than 10 years ago. So now the patterns of RTCs should be even more complicated, because the power of hackers has increased. See, what I mean.
That's right, I also remembered a nuance - but how exactly do all such researchers get SB? (maybe it was somewhere, but I missed it or didn't get to it)
I mean, how much of it is really random.
The biggest poker site in its time gave a description of its RTC engine (it cost money by those standards), so there they used not only high-precision sensors of temperature, pressure in several places of the server room and other similar things, but even data on last mouse movements of users at a particular poker table. I hope this gives an idea of the level of randomness of GSR, which they estimated at four nines (That's all!).
And they came to this not because they wanted to show off their budget. And because more simple models of RNG have practically extremely low resistance to hacking (pattern finding), and there were many precedents of poker rooms' RNG hacking
back in the day. This is more than 10 years ago. So now the patterns of RTCs should be even more complicated, because the power of hackers has increased. You know what I mean.
Yeah, I really don't know what analogy you can make with the market, whether it's perfect or not.
I use, like,https://docs.scipy.org/doc/numpy/reference/random/bit_generators/pcg64.html#r7c40bac0730f-2.
and there turns out a lot of regularities, if you dig
Yeah, I really don't know what analogy you can make with the market, whether it's perfect or not.
I use, for example, thishttps://docs.scipy.org/doc/numpy/reference/random/bit_generators/pcg64.html#r7c40bac0730f-2
and there's a lot of regularities there, if you dig
I.e. my guesses were right, RNG used by mere mortals cannot generate a really random walk anywhere near its mathematical definition.
If on this ground (work with generated SB) the research was built, then this leads to an internal contradiction. It is necessary to look at the problem from a different angle.
i.e. my guesses were justified, the RNG used by mere mortals cannot generate a really random walk any close to its mathematical definition.
If on this ground (work with the generated SB) were built research, then this leads to an internal contradiction. We need to look at the problem from a different angle.
Let's calculate the pattern in conditional raccoons.
For the SB over 20 years:
For the forex euro:
Look for whistles, as they say.
Let's take a smaller one, 10 years:
Better here, check it out in the tester:
Well sort of yes, there is something, but the logic is very simple, by closing prices. And most likely it is not correct, I have just sketched it out.
Recalculated, now it's right:
Hahaha, it's so dumb and simple. At SB it would just be a stick in the sky.
Hahaha, it's so dumb and simple. The SB will just be a stick in the sky.
So, what's the problem? Put your apartment on credit and go to the casino to play - roulette, dice are common SB generators.