Model 3: Regression Models with temperature

(d) Suppose that we think weather impacts ridership. We have collected data on the highest temperature for each month in the dataset. Add the variable 'Highest Temperature' to Model 2.

i. Copy and paste the StatTools report including the regression equation.

Also, display a plot of the fitted values and the actual values over time.

ii. Compute RMSE and MAPE of the model based on the validation data.

iii. Suppose the high temperature is 85 degrees in June 2017. What is your prediction for June, 2017? Show how you got there.

(e) Which model is the best (your exponential smoothing model and two regression models). Explain your answer.

|Month Trips/month Highest Temperature Jun-13 21213 92 Jul-13 31565 98 Aug-13 37009 90 Sep-13 38731 96 Oct-13 36639 86 Nov-13 24390 70 Dec-13 15499 71 Jan-14 10556 58 Feb-14 8834 56 Mar-14 15420 66 Apr-14 24336 77 May-14 30806 86 Jun-14 34523 89 Jul-14 34245 91 Aug-14 34713 90 Sep-14 34440 92 Oct-14 29811 77 Nov-14: 19455 69 Dec-14; 14111 62 Jan-15 10129 56 Feb-15; 7769 43 Mar-15 12057 62 Apr-15 23829 80 May-15 33642 88 Jun-15 33047 90 Jul-15 35960 96 Aug-15 38410 95 Sep-15 42742 97 Oct-15 39106 78 Nov-15 32909 74 Dec-15 25939 72 Jan-16 17306 59 Feb-16 19329 61 Mar-16 29641 79 Apr-16 33769 82 May-16: 39108 92 Jun-16: 48664 88 Jul-16 44512 96 Aug-16 50231 96 Sep-16 54925 91 Oct-16: 50751 85 Nov-16 39890 72 Dec-16: 26194 60 Jan-17 33225 66 Feb-17 28344 70 Mar-17 24712 64 Apr-17 43846 87 May-17 48900 92