M4
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
M4 Hourly
1.004Median MASE
13
M4 hourly 48
0.54Relative MAPE
13
M4
0.036CRPS
13
M4 evaluation set
90MAGE ASR
12
Average M4 benchmark (test)
100AUROC
12
M4
89.33Accuracy
12
m4 Chronos Benchmark II (yearly)
0.091WQL
12
m4 Chronos Benchmark II (quarterly)
0.062WQL
12
M4 (test)
0.03CRPS
12
M4 Yearly
3.147MASE
10
M4 Zero-shot
13.53Yearly sMAPE
10
M4 Others
6.245sMAPE
10
M4 L=168 (test)
0.12MSE
9
M4 L=96 (test)
0.2MSE
9
M4 Monthly (test)
91.1OWA
8
M4 Quarterly (test)
91.6OWA
8
M4 Yearly (test)
0.949OWA
8
M4 Overall
16.369MSIS
8
M4 Others
31.162MSIS
8
M4 Monthly
9.373MSIS
8
M4 Quarterly
10.934MSIS
8
M4 Yearly
33.046MSIS
8
M4 extension Generation: Polish 4o-mini, Regeneration: Para 4o-mini
0.133ROC AUC @ 1% FPR
8
M4 extension Generation Revise 4.1 Regeneration Para 4o-mini
25.26ROC AUC @ 1% FPR
8
M4 extension Generation: Revise 4o-mini, Regeneration: Para 4o-mini
0.414ROC AUC @ 1% FPR
8