T2 - . We want to see if is independently distributed as...

Info iconThis preview shows pages 1–4. Sign up to view the full content.

View Full Document Right Arrow Icon
STA4005B Time Series (2007-2008) Tutorial 2 (22/1, 23/1) (Time & Venue: T9, BMS LT; W5, MMW 704) Example 1 ( AR(1) model) Consider the stationary process , and ) , 0 ( ~ , 2 1 a t t t t WN a a Z Z σ φ µ + + = 1 | | < . Find and for ) ( t Z E ) , ( k t t Z Z Cov + K , 2 , 1 , 0 = k 1
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Estimation of a constant mean: Consider the model: , t t X Z + = µ ) 0 ( = t X E for all time t. where { is } t X stationary with zero mean, variance 0 γ , autocorrelation function (ACF) k ρ . = = = n t t Z n Z 1 1 ˆ is unbiased estimator of .( i.e. = ) ˆ ( E ) Then i) ii) + = = 1 1 0 ) 1 ( 2 1 ) ˆ ( n k k n k n Var iii) The standard error of ˆ is ) ˆ ( Var . Example 2 Suppose ) 1 , 0 ( ~ , 5 . 0 8 . 0 2 1 N a a a a Z t t t t t + = . Find . ) ( 10 1 = t t Z Var 2
Background image of page 2
Example 3 (MA(q) model) Suppose that are independent and identically distributed random variables with mean zero and finite variance and is a constant for If t a i b K , 2 , 1 , 0 ± ± = i , 1 1 0 q t q t t t a b a b a b X + + + = L then compute the autocorrelation function k ρ of { } t X for K , 2 , 1 , 0 = k 3
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Residual Analysis After model is fitted, we need to assess the model by a residual analysis using a graphical approach. Let be the residuals
Background image of page 4
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: . We want to see if is independently distributed as normal with constant variance. t t t Z Z e ˆ − = n t , , 2 , 1 K = t e (a) Whether the residuals look like independent. (Time series plot of the residuals) (b) Whether the residuals look normal. (Histogram or Q-Q plot of the residuals) (c) Whether the residuals look like to have constant variance. (Plot of residuals against the fitted values) The sample autocorrelation function For the stationary process 2 )] )( [( Z k t t k k Z Z E σ µ γ ρ − − = = + The sample autocorrelation function is defined as c c r k k = where , ) )( ( 1 1 ∑ − = + − − = k N t k t t k Z Z Z Z N c K k , , 1 , K = with K not larger than N/4 . is the most satisfactory estimate of k r k . 4...
View Full Document

Page1 / 4

T2 - . We want to see if is independently distributed as...

This preview shows document pages 1 - 4. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online