Scribd
Upload
203 views1 page

ECE3093 Assignment

This document outlines the instructions for Assignment 3, which includes two problems to be completed individually. [1] The first problem asks students to prove whether a given time series is stationary. [2] The second problem has both theoretical and computational components related to simulating and analyzing an ARMA time series process. Students are instructed to submit their work as a single PDF file by the given deadline.

Uploaded by

heinthetzaw
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
203 views1 page

ECE3093 Assignment

This document outlines the instructions for Assignment 3, which includes two problems to be completed individually. [1] The first problem asks students to prove whether a given time series is stationary. [2] The second problem has both theoretical and computational components related to simulating and analyzing an ARMA time series process. Students are instructed to submit their work as a single PDF file by the given deadline.

Uploaded by

heinthetzaw
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 1

ECE3093 Assignment 3

The following problems are to be done individually and uploaded as a single


pdf file, including Matlab code as an appendix. Do not upload Word
or zip files. The first question is theoretical, and can be neatly written by
hand, or typed if you wish. The second question has both theoretical and com-
putational components. You may hand write, or type, the theoretical parts,
but any plots or calculation results should be included and referred to, as well
as the code used to generate the results. The assignment must be uploaded by
11:55 pm on Friday, Week 12 (26 May).

1. Let Xt = A cos(φt + B), where t ∈ Z, A, φ ∈ [0, ∞) are constant, and B


has a continuous uniform distribution on [−π, π] (i.e. B ∼ U ([−π, π])). Is
Xn stationary? Hint: prove each part of the formal definition of station-
arity.(10 Marks)

2. Consider the following ARMA(1,2) time series:

Xn − 0.5Xn−1 = Zn + 0.25Zn−2 ,

with Zn ∼ W N (0, 2).

(a) Calculate the theoretical ACVF using the Yule-Walker algorithm.


(10 Marks)
(b) Simulate 1000 points of the time series and produce an appropriate
plot (see https://au.mathworks.com/help/econ/simulate-stationary-
arma-processes.html for help). (5 Marks)
(c) Write Matlab code to calculate an estimate of the ACVF for 20 lags
using the Matlab autocorr function (https://au.mathworks.com/help
/econ/autocorr.html) or other appropriate function. Plot γ̂(h) (or
ρ̂(h), either is fine) for the points simulated at (b). (5 Marks)
(d) Write Matlab code to simulate 1000 points from the time series and
estimate its ACVF for 20 lags, n times, for n = 1, 10, 100, 1000, 10000, 100000
and 1000000. For each n, average the ACVF estimates and calculate
the Root Mean Square Error (RMSE) with the theoretical ACVF.
Show via suitable
√ plots and analysis that as n increases, the RMSE
reduces as 1/ n. Hint: use a suitable transformation to linearize
the relationship between n and RMSE, followed by regression. (10
Marks)

You might also like