For both approaches, initial
solutions are obtained with a restricted model and lunch breaks are scheduled in a post-processing
step. Experiments were conducted to determine the value of the models and the performance of the
algorithms
using real data provided by MD Anderson Cancer Center in Houston, Texas.