Skip to main content

Table 3 Time periods chosen for data split into training, validation and testing sets

From: Experimental evaluation of baselines for forecasting social media timeseries

Domain

Training (52 days)

Validation (1 week)

Testing (1 week)

Vz19

2018-12-24 to 2019-02-14

2019-02-15 to 2019-02-21

2019-02-22 and 2019-02-28

CPEC

2020-03-30 to 2020-05-21

2020-05-22 to 2020-05-28

2020-05-29 and 2020-06-04

BRIA

2020-02-01 to 2020-03-24

2020-03-25 to 2020-03-31

2020-04-01 and 2020-04-07