Skip to main content

Table 3 MAE rates for all types of hyperparameter tuning predicting hospital demand 1, 3 or 7 days in advance in case of the batch method for St Mary’s hospital

From: A unified machine learning approach to time series forecasting applied to demand at emergency departments

   Choosing the best set of hyperparameters based on:
algorithm days yesterday the past exponential the average over the caret
    n days moving average whole training set  
  1 14.50 14.38 14.48 14.42 14.34
lm 3 14.80 14.96 14.86 14.95 14.80
  7 15.11 15.17 15.22 15.33 15.13
  1 15.97 14.49 15.08 14.46 14.30
gbm 3 15.38 14.63 14.86 14.67 14.53
  7 15.43 14.79 14.93 14.78 14.86
  1 14.49 14.34 14.38 14.35 14.48
glmnet 3 14.81 14.79 14.82 14.75 14.77
  7 14.98 15.11 14.94 15.16 15.09
  1 15.86 15.47 15.74 15.79 15.42
knn 3 16.13 15.78 15.92 16.02 15.57
  7 16.70 16.05 16.45 16.52 15.89
  1 14.62 14.46 14.54 14.79 14.45
rf 3 14.66 14.55 14.54 15.15 14.53
  7 14.93 14.61 14.62 15.53 14.67