Partial autocorrelation function: Is it important from exam point of view? Can anybody help me understanding the following: min E [(something)^2] ???
You just need the formulae given on page 40 of the Tables (which are derived from the expression above)
Theory for Time Series (1) Hi John, The theory for The first-order autoregressive model AR(1) and The autoregressive model AR(p) seems quite quite heavy though I am sort of able to follow the numerical questions. Please help me on how to improve my understanding of the proofs and theory?
The (UK) exams focus on the application rather than the theory - so make sure that you can test stationarity and can derive the autocovaruance, the ACF and the PACF for AR examples - see qns 4.4, 4.6, 4.7 and 4.10 in the Q&A bank and X4.13 in the assignments. If you want you can drop me an email and I can send you the old C2 time series questions which are pretty much all calculate the ACF and make good practice.
Hi John, Regarding chapter Time Series 1 and Auto Correlation Function and Partial Auto Correlation Function, I am unable to understand how both are equal at lag 1 as it’s mention on Page 18 question and Page 19 information. It appears that I am missing something. Could you please help. Thanks. Sunil
The PACF at lag 1 is the conditional correlation between, say X1 and X2, given all the values in between (ie fixing those values). Since there are no values between X's 1 apart, it will be the same as the ACF.