Develop a forecast for both unemployment rates for January (Boston) and February (United States) 2024. You must test at least three different forecasting methods (no more than two iterations of a same method (e.g. you can do 3 month and 6 month Moving Average, but you must test at least one other method). Pick the one you think is the best, and explain why you picked it. Provide the forecast unemployment rate for February for U.S. and January for Boston. Does the forecast "make sense" to you?
At the very least you should answer the following questions:
1. Which criteria you used to choose the forecast (e.g. RMSE, MAD, etc.)? How did different models perform on this criteria (i.e. report actual numbers)? What was the forecast of each model?
2. What is your final forecast based on the model(s)?
3. Do you think it is an under-estimate, over-estimate or fairly reasonable? If you think the rate would be something other than what your method produces, what is your predicted unemployment rate for each? Explain your choice.
UNGRADED BUT SEPARATELY REWARDED PORTION:
What is your final forecast (using the model(s) and/or your subjective judgment) to two decimal points (e.g. 3.11) for February (Boston) and March (US) (i.e. one month later than your forecast above) ? Most accurate forecast (not going over the actual rate in a case of a tie) will receive a prize once the actual February (Boston) /March (US)* unemployment if announced by the Bureau of Labor Statistics.
