Human resource planning with worker attendance uncertainty

We consider a human resource supply planning problem given a set of job demands. The staffing levels of different workforce types to fulfill the jobs need to be determined, a priori to full knowledge of the attendance rates of the workers. The objective is to find the optimal staffing levels that minimize the hiring costs while maintaining a high certainty of fulfilling all the jobs. We propose six different approaches for generating solutions to the problem. Computational experiments are conducted to evaluate the performance of the approaches. Our experimental results show that two approaches: one based on stochastic programming and the other based on robust optimization using an ellipsoid uncertainty set, outperforms the other approaches consistently in various performance measures.