Read Section 10.3, “Non-stationary seasonal ARIMA models,” on pages 233-234.

 

Read Section 10.4, “Model specification, fitting, and checking,” on pages 234-241. These are illustrations; they help you with your student project, but they no material to know for the final exam.

 

Read Section 10.5, “Forecasting seasonal ARIMA models,” on pages 241-245. These are examples; you need not memorize the equations. Any exam problems are structured to be answered intuitively, not by equation.

 

Many time series are seasonal. Check if the time series in your student project is seasonal. This module help you understand how seasonality affects the observed time series values.

 

NEAS:

I want to know what are the different seasonalities signified by these two seasonality model:
Y(t)=mu(t)+et which is in chapter 3;
and MA process,
Y(t)=e(t)-theta*e(t-12) which is used to model seasonality in chapter 10
I can't figure out,
is it true that first one means yt is expected as mu which is deterministic for a certain month and taht second one means the seasonality of yt depend on what happened in the same month last year and it is stochastic?

[NEAS: They are both stochastic. Consider two clearly seasonal time series: daily temperature and sales at a sports store. The average daily temperature in August in 20X1 is best modeled as the average daily temperature for August plus a stochastic factor. The average daily temperature in August 20X0 is not relevant. The sports store has sales for each sport that depend on the season. Skiing gear has high sales in winter; swimming gear has high sales in summer. But sales are best forecast based on the previous year’s sales in the same month. If August 0X0 had high sales of swimming gear, August 20X1 may have high sales as well.]

 

I want to know how would one model the seasonality of auto insurance?Can any one help me with this?


Wedding venues