When measuring productivity of a department it is important to consider absenteeism. Whilst it is expected that some employees will be absent from work, it is necessary to know if a particular absence will likely lead to an extended time off work.
Once predictions are made it will be possible to organise the rota in a more efficient way and distribute the workload more evenly.
The model splits absenteeism into four main categories:
- Various types of diseases
- Pregnancy related
- Poisoning or unclassified reasons
- ‘Light’ reasons such as dental appointments
Using this along with other inputs the model will predict, using logistic regression, how whether a specific absence will be an extended absence.