Bayesian regression - Has anyone made an EA using this algorithm? - page 24
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
For example. There are 30 predictors - inputs. From these, we need to select a subset of predictors that have the strongest impact on the target variable. In my examples it selects 10-15 predictors that are relevant to the target. Gives statistics with confidence intervals.
They are optimisable parameters. Although you'll say I'm not optimising parameters, I'm looking for confidence intervals... :)
Regarding the number of predictors I found an interesting thing on the net:
https://www.mql5.com/ru/forum/71817
Mathematicians formulate it in terms of approximation of arbitrary nonlinear function by broken line, i.e. with increasing number of independent variables one can approximate broken line to any curve as precisely as he wants, but this broken line is not good for anything else.
I also read that by enumerating predictor values one can reach local maxima of a function of unknown form - profit=unknown function(predictors) - on a given sample, but this does not apply to other samples that do not overlap with the first.
And "TC's explanation on the stamp"
For example. There are 30 predictors - inputs. From these, we need to select a subset of predictors that have the strongest impact on the target variable. In my examples it selects 10-15 predictors that are relevant to the target. Gives statistics withconfidence intervals.
Don't neural networks do the same thing?
Is it possible to make a million dollars on the forex market? If so, who will pay it out? The company we trade through? Where does it get that kind of money? From clients who pay spreads and lose, that is, by the definition of the Forex market the office loses when its clients win. If all clients win, where does the money come from to pay them all? Insurance? So forex brokers are insurance companies. Clients pay for insurance with spreads and commissions in the hope that these payments will come back to them with 400:1 leverage. But then again, if everyone loses, the broker goes bankrupt. Then back to the original question: is it possible to make a million dollars on forex? Probably not, because it is more profitable for the office to close and appear under another name. But this is a fraud! They have to be prosecuted! Pardon me, what law? The office is opened in a place where the law doesn't exist. How do we prosecute? Do you understand the logic of your thinking?
What is this?
It's a famous statement by a Western guru about the simplicity of TC. You don't know that either?
Unfortunately, no. And I don't trust gurus, including Western ones. :)
That's right. You can't be trusted. And certainly not the Western ones.
That's why you have to open books and come to an understanding with your own mind. And not in a "summary".
NS is one of the machine learning algorithms, I don't use them.
But I know about the other randomforest algorithm: there are variants of it that use the above algorithms in training.
That's why you have to open books and come to an understanding with your own mind. And not in a "summary".
NS is one of the machine learning algorithms, I don't use them.
But I know about the other random forest algorithm: there are variants of it that use the above algorithms in training.