Skip to main content
. 2021 Apr 22;16(4):e0250417. doi: 10.1371/journal.pone.0250417

Table 3. Improvement percentage of our predictive performance compared with LSTM in terms of MAE in the five most and least improved prefectures.

Lower values indicate greater improvement because a lower MAE indicates better performance.

Rank 2016/31st–2017/30th 2017/31st–2018/30th 2018/31st–2019/30th 2019/31st–2020/30th
Prefecture Improve ment (%) Prefecture Improve ment (%) Prefecture Improve ment (%) Prefecture Improve ment (%)
1 Tokushima -79.5 Aomori -46.1 Oita -50.7 Kochi -61.1
2 Kagawa -75.0 Nigata -45.5 Gunma -48.2 Kagoshima -60.5
3 Hiroshima -74.0 Fukui -39.9 Okayama -47.8 Wakayama -60.4
4 Okayama -69.8 Ishikawa -39.6 Ehime -47.6 Miyazaki -58.7
5 Yamaguchi -66.9 Toyama -38.7 Kagawa -47.2 Saga -55.7
43 Gifu -34.0 Okinawa -14.9 Shizuoka -17.3 Akita 9.1
44 Shiga -32.3 Kyoto -12.5 Tokyo -14.5 Hukushima 10.6
45 Fukushima -27.8 Kochi -11.6 Okinawa -13.1 Nagano 14.4
46 Yamagata -24.2 Okayama -11.5 Yamaguchi -5.0 Aomori 15.9
47 Okinawa -17.5 Shiga -9.1 Miyazaki 1.6 Hokkaido 20.5