Case 2. Forecasting Emergency Department (ED) Visits (30 marks) The spreadsheet contains data on the number of visits to the Emergency Department (ED) of a hospital in our region for the last three weeks. The data excludes arrivals by ambulance and provides breakdown of the total number of patient visits by day of the week and 6-hours shifts. Based on this last three weeks of data, the ED manager is seeking help to forecast ED demand in terms of the number of patient visits for each day of the week and shifts over the next week. To help the ED manager achieve his goal, perform the following analyses, and assume the effect of seasonality to be a multiplicative effect. a) Explore the data, create graphs and comment on the seasonality patterns you observe in the data. Explore the seasonality patterns by \"day of the week\" as well as by \"day of the week and shifts." (2 marks) b) Use the \"seasonality forecasting technique with no trend\" to forecast the volume of ED visits by \"day of the week" only. (Use the actaat1 number of ED visits as the initial vatnefor exponential smoothing. Use smoothing constant 0.25) (8 marks) 0) Calculate MAD and MAPE measures of error on training data for the forecasting model in part Cb). (3 marks) d) Plot the actual volume of ED visits versus forecasted volume from part (b) and comment on how they compare. (1 mark) Shift Week Day of Week 12 AM - 06 AM 06 AM - 12 PM 12 PM - 06 PM 06 PM - 12 AM G UI A W N Monday 53 74 85 53 Tuesday 52 70 86 52 Wednesday 46 66 87 56 Thursday 52 73 84 52 8 Friday 53 77 82 53 9 Saturday 46 62 70 42 10 Sunday 51 77 77 51 11 Monday 58 81 93 58 12 Tuesday 45 70 80 55 13 Wednesday 44 67 67 44 W W W W N N N N N N N P PP PP 14 Thursday 48 62 76 44 15 Friday 51 79 96 56 16 Saturday 42 59 68 42 17 Sunday 47 60 69 39 18 Monday 56 78 81 64 19 Tuesday 52 76 84 50 20 Wednesday 46 64 74 46 21 Thursday 46 65 72 58 22 3 Friday 54 76 87 54 23 W Saturday 47 65 74 38 24 Sunday 46 64 73 46