The M4 time series competition kicked off earlier this month and already features teams from 37 countries across the globe, with many participants hailing from the USA, India, the UK and Brazil. This latest edition of the M-Competition is on track to being the largest of its kind, and remains open to all avid forecasters.
The most popular type of forecasting method currently used by participants is a “Combination” of forecasts, followed by “ML” methods and “Statistical” methods at about the same level of utilization.
Prize money for the M4 currently amounts to €25,000, which is set to increase as additional sponsors come on board. The global transportation technology company, UBER, is the latest organisation to join, and will be awarding a Student Prize for the Most Accurate Forecasting Method.
The deadline for submissions is 31 May 2018, which allows forecasters ample time to submit their forecasts and prediction intervals. We do wish to remind you that there are 100,000 time series, covering various types of areas and different time frequencies.
Complete information about the M4 Competition can be found on the M4 website. Additional information can also be found in the user guide.
| Frequency | Demographic | Finance | Industry | Macro | Micro | Other | Total |
|---|---|---|---|---|---|---|---|
| Yearly | 1,088 | 6,519 | 3,716 | 3,903 | 6,538 | 1,236 | 23,000 |
| Quarterly | 1,858 | 5,305 | 4,637 | 5,315 | 6,020 | 865 | 24,000 |
| Monthly | 5,728 | 10,987 | 10,017 | 10,016 | 10,975 | 277 | 48,000 |
| Weekly | 24 | 164 | 6 | 41 | 112 | 12 | 359 |
| Daily | 10 | 1,559 | 422 | 127 | 1,476 | 633 | 4,227 |
| Hourly | 0 | 0 | 0 | 0 | 0 | 414 | 414 |
| Total | 8,708 | 24,534 | 18,798 | 19,402 | 25,121 | 3,437 | 100,000 |
