1. Source: Time Series Data Library www.datamarket.com MONTHLY MINNEAPOLIS PUBLIC DRUNKENNESS INTAKES JAN.’66-JUL’78 Meghan Burke

2. GOALS  Provide a descriptive analysis of this time series  Develop the best models to forecast future monthly Minneapolis public drunkenness intakes  Understand the time series structure

3. INITIAL TIME SERIES PLOT

4. TIME SERIES COMPARISON Initial Adjusted

5. TIME SERIES COMPARISON Initial Adjusted

6. MODEL COMPARISONS  Best Smoothing Models:  Seasonal Exponential Smoothing  Winters Method  ARIMA (0, 0, 1)

7. SEASONAL EXPONENTIAL SMOOTHING

8. WINTERS METHOD

9. ARIMA (0,0,3)

10. FORECAST PREDICTIONS FROM THE EXPONENTIAL SMOOTHING MODELS

11. COMPARING FORECAST PREDICTIONS  Best Smoothing Model  Seasonal Exponential Smoothing ModelMAPEMAE Seasonal Exponential Smoothing 14.13256252%38.52444229 Winters Method 14.13256243%38.52444203 ARIMA (0,0,3) 21.72210497%60.33186319

12. CITATION  David E. Aaronson, C. Thomas Dienes, and Michael C. Musheno,Changing the Public Drunkenness Laws: The Impact of Decriminalization, 12Law & Soc'y Rev.405 (1977), Available at: http://scholarship.law.berkeley.edu/facpubs/113