Discussion of article "R-squared as an estimation of quality of the strategy balance curve" - 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
Don't mould as we know how - at least for ourselves.
p.s. You have read diagonally. R^2 is calculated not for quotation charts, but for equity. Its main task is to show even equity. That's it. Whether it corresponds to random or not - let the user decide that
It doesn't matter what time series: quotes or balance. You just need to use normal software and understand WHAT this normal software produces. And this software must provide information that says: "Is it possible to use it at all?".
Why don't you take the balance and fit a NORMAL linear regression (in R it is ln). It may well be that for the balance we will get significant coefficients and then everything you have written has a right to life, but if not, then sorry....
The article is about a custom criterion for obtaining smooth equity curves. The proposed criterion and implementation fully achieves the goal.
You decided the article was about something else and started refuting what you thought it was about. Five for fantasy, two for reading.
The article is about a custom criterion for obtaining smooth equity curves. The proposed criterion and implementation fully achieves the goal.
You decided the article was about something else and started refuting what you thought it was about. For fantasy, five, for reading, two.
Ours doesn't understand yours.
Or rather you don't understand statistics, since I wrote that your R-squared may not even exist. Any figure you get in statistics, you have to prove that it exists.
This is actually the third post about the same thing.
Don't understand, don't want to understand - good luck.
I was writing because I hoped to be useful.
More precisely, you don't understand statistics, because I wrote that your R-squared may not exist at all. Any figure that you get in statistics, you have to prove that it exists.
The article has nothing to do with Statistics! You have invented an image of the article and are fighting this mill.
SanSanych, don't interfere. We mould as we know how!
Don't mould as we can - at least for ourselves.
SanSanych, we understand your point of view. Indeed R^2 does not have an additional parameter that evaluates its validity. However, I would like to point out that none of the MetaTrader standard report statistics has such a parameter. What is the probability of reliability of Net Profit, Take Profit, Sharpe Ratio, etc.? What is the confidence level of all these numbers? There is no such parameter, but this does not prevent you from using the report.
And by the way, this is a great idea. The parameter of reliability of the obtained results is necessary. For example, we have received a tester's report with a positive result and good statistics. What is the probability that this result is random? How to estimate this probability? How to calculate it? How to implement it into a standard report? These are questions you can answer and show others how it can work. You could make a really valuable contribution to the community by writing a relevant article. I can't write such an article - I don't have enough knowledge, but you can, so you are in higher demand.
SanSanych, we understand your point of view. Indeed, R^2 does not have an additional parameter that evaluates its reliability. However, I would like to point out that none of the MetaTrader standard report statistics has such a parameter. What is the probability of reliability of Net Profit, Take Profit, Sharpe Ratio, etc.? What is the confidence level of all these numbers? There is no such parameter, but this does not prevent you from using the report....
SanSanych will expose you rascals :-))))
Vasily, how can the indicator itself (for example, R^2) have a parameter that evaluates its reliability? That is what the chosen statistical method does.
In the article, the author has 3.3 Correlation Testing. Hypotheses, significance level and all that....
fxsaber:
The article has nothing to do with Statistics! You have invented an image of the article and are fighting with this mill.
SanSanych will expose you rascals :-))))
Vasily, how can the indicator itself (for example, R^2) have a parameter that evaluates its reliability? That's what the chosen statistical method does.
That's funny. I was still laughing yesterday...Don't get hung up on words. Of course I meant what you wrote.
That's funny. Yesterday I was still laughing...
Seriously, fxsaber is right. The article is about how to choose an even equity. It doesn't say a word about the fact that all your even equity may turn out to be a wild fit or just luck.
http://studopedia.org/8-161613.html
Dear connoisseurs, please tell me how to correctly attach the estimation of multiple regression coefficients to alglib for its greater usefulness :)
Great article. Thank you for posting this.