The annual revenue data for Amazon spanning 2005 to 2020 is provided:

Period	Year	Revenue (Millions USD)
1	2005	$8,490
2	2006	$10,711
3	2007	$14,835
4	2008	$19,166
5	2009	$24,509
6	2010	$34,204
7	2011	$48,077
8	2012	$61,093
9	2013	$74,452
10	2014	$88,988
11	2015	$107,006
12	2016	$135,987
13	2017	$177,866
14	2018	$232,887
15	2019	$280,522
16	2020	$386,064

A) Make a duplicate copy of the worksheet in Minitab. Delete the final entry for 2020, in the duplicate worksheet. Develop a time-series plot (STAT > Time series > Time series plot or GRAPH > TIME SERIES PLOT). Edit the x-axis by:  double-clicking the graph, in the Edit Graph window, right-click the x-axis and then: (1) under the TIME tab Time Scale > Calendar > Year and for Start Value enter 2005, click OK. Close the edit window. Include your graph here:

B) Using your results from part-a and Minitab, develop a forecasting model. Go to STAT > Time Series > Trend Analysis. Pick the appropriate form of the trend and include the graph output here:

C) Insert a column and use CALC to create a column of values that are a transformation of the Revenue values. (Use your results from the previous parts to ascertain the most appropriate transformation.)  Repeat part b using the transformed values. Insert your graph here:

D) Using the estimated forecast equations from parts a and b (these are provided at the top of each graph) and showing hand calculations (that is, state the equation(s) you use and how you obtain the result from that), forecast the next period (i.e. the year 2020 which we had omitted from this data set). Ensure that all results are in Millions USD (as was the original data) and compare these two results with the actual revenues for 2020. Summarize your findings and comment.

E) Using the annual data and Minitab, run a regression model using Period as the predictor variable and Revenue as the response. Store the residuals, request a 4-in-one plot of the standardized residuals, and request a Durbin-Watson statistic. Insert your 4-in-one plot, Durbin-Watson statistic output and your test of autocorrelation here (make sure to include a statement of the hypotheses, decision rule, and conclusion):

F) Using Excel or Minitab and your stored residuals from the last part, determine the correlation coefficient of the residuals in consecutive periods (i.e. for e_t and e_t-1). Provide your numerical answer and explanation of the result.

G) Using the annual revenue data and Minitab, develop a regression model using the ln(Rev) as the response and period and lag-Residuals as the two predictors. Evaluate your model (coefficient of determination, overall suitability, individual predictors, Durbin-Watson) and predict (showing hand calculations) the revenues for 2020 (include a statement). Compare your result to the actual reported annual revenues. Include all pertinent output.