A multi-neighborhood and multi-operator strategy for the uncapacitated exam proximity problem

A strategy featuring multiple local search operators and multiple neighborhood structures is applied to the exam proximity problem. The use of multiplicity is to enable effective interplay between intensification and diversification during the search process. The algorithmic design is inspired by the variable neighborhood descent algorithm and the "one operator, one landscape" point of view. Its performance was evaluated using publicly available datasets. For the uncapacitated exam proximity problem it was able to produce the best proximity costs for several datasets.