Bayesian regression - Has anyone made an EA using this algorithm? - page 13
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 answered your first question. I really don't understand about the signs. Find the number of bars at which the theory works? I reject it at once.
"The original aim was to reconcile the straight line and the price series. - if the Bayesian regression is a straight line, then it is really no good.
If it is compatible with a straight line, the least squares linear regression (LOS) known to all is enough. Also by ANC method it is possible to combine with any curvilinear. In known to all code instead of number 1,2,3... the values of the curvuline are used.
There can even be a curvuline of unknown form (polynomial) - polynomial regression, the codebase has a code for that.
---
Signs. This is the basis of Bayesian regression. Traits are defined, the presence of which assigns a sample to a particular class with some probability. Having several features and their probabilities, the final probability is calculated using the Bayes formula.
"The sum of a sufficiently large number of weakly dependent random variables of approximately the same magnitude (none of the sums is dominant or determinative) has a distribution that is close to a normal distribution" (Wikipedia).
What makes you think that? Not at all. You don't have to think about it, it's like defining the scope of the Bayesian regression.
We need to determine the features that are needed to calculate the Bayesian regression. This is the first question of how to make a square circle. This is where you may realize that the Bayesian regression does not fit in at all. But we don't care... something has to be done. Suppose that the coincidence of price values of one row and the second row (in our case the line) will correspond to the maximum likelihood. And the maximal one by one path will be 1/n (n - number of bars). Although this approach is just like drawing with a pitchfork in water. So we should invent some formula which at argument 0 gives 1/n, and at increasing argument tends to 0. Then we write down the baes formula and substitute the formula we invented earlier for the probabilities. Next we need to find the maximum of the resulting function. Probably take the derivative, equate it to zero...
The result will be almost the same as the linear regression, because the original purpose was to combine the straight line and the price series.
Having read a bit of literature, it becomes clear that in the Bayesian regression, the estimate of the linear regression coefficients is based on the a priori knowledge of their distribution and assumption of normality of errors. Everything else is the same as in the usual linear regression with ANC estimation of the coefficients. Whether or not to apply it to the market is up to you.
After reading a little bit of math, it becomes clear that in Bayesian regression, estimation of linear regression coefficients is based on a priori knowledge of their distribution and assumption of normality of errors. Everything else is the same as in the usual linear regression with ANC estimation of the coefficients. Whether or not to apply it to the market is up to you.
...
"The sum of a sufficiently large number of weakly dependent random variables, having approximately the same magnitude (no single summand dominates, no determinative contribution to the sum), has a distribution that is close to normal."(Wikipedia)
...
Where did this come from?
English-language Wiki articles and a couple of lectures on the subject. MNC is replaced by Bayesian inference with likelihood maximisation.
And I think you've confused the application of Bayesian theorem to the posterior estimate of the probability of an event occurring with what is done in Bayesian regression. Although both are based on a Bayesian approach to probability.
English-language Wiki articles and a couple of lectures on the subject. MNC is replaced by Bayesian inference with likelihood maximisation.
And I think you've confused the application of Bayesian theorem to the posterior estimate of the probability of an event occurring with what is done in Bayesian regression. Although both are based on a Bayesian approach to probability.
And what and how is there any confusion here?
What is the likelihood?
... forex data has a normal distribution, and hence is the domain of Bayesian regression ...
During some periods "forex data" (let us assume that it is prices) may have a normal distribution, but this is obviously not the case with the trend - perhaps there is a mixture of normal(?) and other distributions.
We can assume that in the price series there is a sequential change of distributions (or their mixtures), not necessarily normal ones.
Applying any regression to price series makes no sense because price series are non-stationary. In Russian, this means that the regression coefficients calculated on one sample will not match those on another sample.
This 18 doesn't cover anything. It's perfectly replaceable by linear regression and the Fibo level. You can't have a normal conversation, you don't have any constructive conversations. You haven't even demonstrated yet that you understand what the 18 is and what it does.
Assess the power of (18) with a simple example, data from here http://www.statdata.ru/russia, what regression can replicate such a thing? You can plug in all 10 top regression methods http://datareview.info/article/10-tipov-regressii-kakoy-vyibrat/
...what regression can replicate something like this? You can plug in all 10 top regression methods http://datareview.info/article/10-tipov-regressii-kakoy-vyibrat/
...