Timeline for How to estimate real series from smoothed moving average?
Current License: CC BY-SA 3.0
11 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jan 18, 2016 at 7:31 | history | edited | mpiktas | CC BY-SA 3.0 |
added 24 characters in body
|
| Jan 16, 2016 at 20:10 | comment | added | mpiktas | Yes cf is coef(mod), this is a typo. | |
| Jan 16, 2016 at 18:40 | comment | added | EngrStudent | Also, if you are going to think of the error, you might look at the majority, not the first 5 terms. You might also consider some "typical magnitude" scaling, like the median absolute value of "x". This is like a relative error, but does not suffer from the problem of very small "true" values and small errors exploding. | |
| Jan 16, 2016 at 18:32 | comment | added | EngrStudent | your variable "cf" is not assigned in code. Do you mean to have something like " cf <- mod$coef ". | |
| Jan 16, 2016 at 18:31 | comment | added | EngrStudent | I think a qqnorm plot might be more informative than a trendplot. | |
| Jan 16, 2016 at 3:38 | comment | added | Aksakal | @mpiktas, it depends on the data. Some series are known to be dependent from prior experience or understanding the process. For instance, it's simply not reasonable to assume that deposit balances or area temperatures are independent or even not autocorrelated. I don't need to test this. Generally, you can't test for independence even if you observe the data. | |
| Jan 15, 2016 at 23:25 | comment | added | mpiktas | This approach can be generalised for arima type $x_t$, but then recovering the original $x_t$ might be difficult, without the additional restrictions. Furthemore even if $x_t$ are not independent, how would you test it, if they are not observed? | |
| Jan 15, 2016 at 22:54 | comment | added | Aksakal | This assumes that original series are not correlated, which may not hold depending on the data. For instance, in economic data this almost never holds, series like sales or customer visits are correlated. | |
| Jan 15, 2016 at 22:45 | vote | accept | Ant | ||
| Jan 15, 2016 at 22:45 | comment | added | Ant | Ah, I see.. thank you, this helps. Initially I thought I could not apply directly an MA estimation because I thought the $x_t$ needed to be independent, but I take it that being uncorrelated is enough? :) | |
| Jan 15, 2016 at 15:48 | history | answered | mpiktas | CC BY-SA 3.0 |