Explore BrainMass
Share

Time Series Models

This content was STOLEN from BrainMass.com - View the original, and get the already-completed solution here!

Is it true that one of the important properties of time series models is that some sort of pattern in the data over time should be detected first and then used within the model? Please clarify.

© BrainMass Inc. brainmass.com October 24, 2018, 6:29 pm ad1c9bdddf
https://brainmass.com/statistics/time-series-analysis/time-series-models-42053

Solution Preview

Please see response attached, which is also presented below. I hope this helps and take care.

RESPONSE:

1. Is it true that one of the important properties of time series models is that some sort of pattern in the data over time should be detected first and then used within the model? Please clarify.

Sometimes that is the case, but not all of the time. For example, if a pattern is observed in the number of car accidents on a certain highway or intersection (i.e., more accidents seem to occur during rush hour), then a researcher may decide to observe that highway or intersection using a time series model. In other words, ...

Solution Summary

In relation to time series models, this solution explains if it is true that one of the important properties of time series models is that some sort of pattern in the data over time should be detected first and then used within the model. The explanation includes an excellent definition including the properties of time series models, and several illustrative examples to explore the truth of the above proposition.

$2.19
See Also This Related BrainMass Solution

Time Series: seasonality, trend-seasonal model, seasonal indexes

Question 1 (30 pts.): The Excel file attendance.xls contains the daily attendance data at a theme park for a period of four weeks in summer. The park manager wants to use these data to make forecasts for the following summer week.

a) Make a time-series plot of the data. Is there a trend? Is there seasonality in the data? What is the seasonality period? (5 pts.)
b) Fit a trend-seasonal model to the data and make forecasts for the 7 days of the next week. Do the forecasts appear reasonable? (5 pts.)
c) Fit only a seasonal model to the data and make forecasts for the next week. Do the forecasts appear reasonable? Which model would you choose? Answer based on the reasonableness of the forecasts and the MAPE values for the two models. (8 pts.)
d) Interpret the seasonal indexes for Saturday and Sunday from part (c) in words. Also explain what the deseasonalized (seasonally adjusted) series means. In particular, explain why in Week 2, although the Saturday attendance (12,714) was higher than the Sunday attendance (11,977), actually the theme park did better on Sunday than on Saturday. (12 pts.)

Week Day Attendance
1 Mon. 4392
1 Tue. 4428
1 Wed. 5143
1 Thu. 6510
1 Fri. 7175
1 Sat. 13796
1 Sun. 11791
2 Mon. 4440
2 Tue. 5072
2 Wed. 5553
2 Thu. 6250
2 Fri. 7119
2 Sat. 12714
2 Sun. 11977
3 Mon. 4523
3 Tue. 5021
3 Wed. 5530
3 Thu. 6279
3 Fri. 7105
3 Sat. 13116
3 Sun. 11982
4 Mon. 3594
4 Tue. 4364
4 Wed. 5732
4 Thu. 6648
4 Fri. 7795
4 Sat. 12937
4 Sun. 11425

View Full Posting Details