辅导ECMT3130留学生、讲解R程序、讲解natural logarithms、辅导R
讲解SPSS|解析Haskell程序
ECMT3130: Forecasting for Economics and Business
Group Project with Individual Component?
For this project, you should work—individually and in groups—with the
monthly time series of the assigned commodity prices. The project consists
of a set of time series estimation and forecasting exercises, as outlined below.
You should address all these items. You should upload the reports and the
R file containing all the coding for your project by the due dates. Questions
1–4 belong to the Individual Component of the assignment. You should
address these by the due date of the mid-semester test. The remaining
questions belong to the Group Project. You should address these by the
group project submission date.
1. Transform the prices to their natural logarithms, and work with these
transformed series, denoted by pt
, henceforth.
2. Divide the series into the in-sample estimation environment (observations
up to and including Dec 2010) and the out-of-sample forecasting
environment (observations beginning from Jan 2011 and onward).
3. Using data from the estimation environment, obtain and report parameter
estimates of the following models:
(a) polynomial deterministic trend model with seasonal component
(use AIC to decide on the polynomial order);
(b) autoregressive model of order p (use AIC to decide on the autoregressive
order)
(c) autoregressive model of order p with linear trend and seasonal
component (use AIC to decide on the autoregressive order).
4. For the same set of models as in the previous question, as well as the
random walk model (i.e., four models in total), obtain and plot point
forecasts for the duration of the out-of-sample window size (i.e., for
periods from January 2011 to December 2017).
The individual component will be assessed as part of the mid-semester test
15. Based on the category of commodities, prices of which have been assigned
to your group, briefly (no more than 500 words) motivate the
topic (as if you were preparing a research report). This will serve as
an ”introduction” of your report.
6. Present a table with descriptive statistics (that lists mean, standard
deviation, minimum and maximum values of the times series), and
another table with unit root test results for the assigned commodity
prices. Briefly (no more than 600 words) describe key characteristics
of the time series. This will serve as a ”data” section of your report.
7. For all four models in Question 4, using a rolling window approach:
(a) obtain one-step ahead forecast errors, and calculate the out-ofsample
root mean square forecast error (RMSFE) measures;
(b) test the hypotheses of (i) error unbiasedness; (ii) error efficiency;
and (iii) no autocorrelation.
(c) test the hypotheses of equal forecast accuracy of all models relative
to the random walk model.
8. Present a table with these results. Briefly (no more than 1000 words)
describe key findings of your analysis. This will serve as a ”results”
section of your report.