1.
Schematic overview of the MILP-based schedule construction and validation phases before implementation.
Data used as model inputs for the computational experiments of each center.
Key performance indicators defined by the RT centers.
Main infeasibilities and limitations found during the schedule construction method and corresponding model adjustments.
Performance comparison between the schedule built manually and the MILP model in the NKI.
Performance comparison between the schedule constructed manually and the MILP model in the BVI.
Data imputation for 2015–2017.
Data imputation for 2017–2019.
Correlations of value-added scores.
Covariance table of VA scores with different outcome domains (mathematics and language) over time.
