BA 555 group project #3
BA 555 Practical Business Analysis
Group Project 3: Time Series Analysis and Forecasting
Due: March 14, 2013 at the beginning of the class
NAME
NAME
NAME
1. Insert a time series plot. Comment on the underlying trend and seasonal patterns. This is your own observation. There is no need to run any forecasting model here.
(Insert the plot here.)
(Insert your comments here.)
2. Forecasting using a Multiplicative Model:
a. Use the time series decomposition method (textbook Chapter 18.6; packet pp. 78-80) to deseasonalize the time series and obtain seasonal indexes. Fit a linear trend model to the deseasonalized time series. Report the fitted model and the seasonal indexes.
Seasonal Pattern
Month
Seasonal Index
January
1.154008
February
1.00262
March
1.087917
April
0.923024
May
1.058668
June
0.882363
July
0.997589
August
1.118552
September
0.826317
October
0.888524
November
0.908866
December
1.039627
Linear Trend
b. Describe the underlying trend and seasonal patterns based on the multiplicative model developed in Question 2-a. Quantify your statements. For example, “There is an upward trend.” and “Sales in December is higher.” are not acceptable and will result in point deduction. “Sales goes up by about $20,000 every month, give or take $1,000 for a 95% confidence.” and “Sales in December is usually 16% higher than the average sales in a typical year.” are what I am looking for. Does this model support your observation in Question 1?
Sales goes up by about $183,010 every month, give or take $262 for a 95% confidenc
Sales in December is usually 16% higher than the average sales in a typical year;
c. Use this model and forecast sales for January through December of the fourth year. Report the forecasts in the table below.
Sales Forecasts
Month
Sales Forecasts
January
Forecast
February
March
286.75
April
267.08
May
272.42
June
216.42
July
Group Project 3: Time Series Analysis and Forecasting
Due: March 14, 2013 at the beginning of the class
NAME
NAME
NAME
1. Insert a time series plot. Comment on the underlying trend and seasonal patterns. This is your own observation. There is no need to run any forecasting model here.
(Insert the plot here.)
(Insert your comments here.)
2. Forecasting using a Multiplicative Model:
a. Use the time series decomposition method (textbook Chapter 18.6; packet pp. 78-80) to deseasonalize the time series and obtain seasonal indexes. Fit a linear trend model to the deseasonalized time series. Report the fitted model and the seasonal indexes.
Seasonal Pattern
Month
Seasonal Index
January
1.154008
February
1.00262
March
1.087917
April
0.923024
May
1.058668
June
0.882363
July
0.997589
August
1.118552
September
0.826317
October
0.888524
November
0.908866
December
1.039627
Linear Trend
b. Describe the underlying trend and seasonal patterns based on the multiplicative model developed in Question 2-a. Quantify your statements. For example, “There is an upward trend.” and “Sales in December is higher.” are not acceptable and will result in point deduction. “Sales goes up by about $20,000 every month, give or take $1,000 for a 95% confidence.” and “Sales in December is usually 16% higher than the average sales in a typical year.” are what I am looking for. Does this model support your observation in Question 1?
Sales goes up by about $183,010 every month, give or take $262 for a 95% confidenc
Sales in December is usually 16% higher than the average sales in a typical year;
c. Use this model and forecast sales for January through December of the fourth year. Report the forecasts in the table below.
Sales Forecasts
Month
Sales Forecasts
January
Forecast
February
March
286.75
April
267.08
May
272.42
June
216.42
July