figshare
Browse
pone.0296878.g003.tif (646.28 kB)

Fig 3 -

Download (646.28 kB)
figure
posted on 2024-02-02, 18:29 authored by Ahmad Kakakhail, Aimal Rextin, Jeroen Buters, Chun Lin, José M. Maya-Manzano, Mehwish Nasim, Jose Oteros, Antonio Picornell, Hillary Pinnock, Jurgen Schwarze, Osman Yusuf

Mean Absolute Error (MAE) calculated over various time intervals using their corresponding meteorological data for the start day prediction under the 2.5% and 5% start day criteria: (A) the various starting points (B) the various ending points. The number of total combinations considered in this brute force optimization was 11 × 56. Only the combinations of starting points yielding the best ending points (and vice versa) are shown. This was only done for the training years as the testing years were reserved to evaluate the performance of the algorithm.

History