Wim Vancroonenburg

One hyper-heuristic approach to two timetabling problems in health care

We present one general high-level hyper-heuristic approach for addressing two timetabling problems in the health care domain: the patient admission scheduling problem and the nurse rostering problem. The complex combinatorial problem of patient admission scheduling has only recently been introduced to the research community. In addition to the instance that was introduced on this occasion, we present a new set of benchmark instances. Nurse rostering, on the other hand, is a well studied operations research problem in health care. Over the last years, a number of problem definitions and their corresponding benchmark instances have been introduced. Recently, a new nurse rostering problem description and datasets were introduced during the first Nurse Rostering Competition. In the present paper, we focus on this nurse rostering problem description.The main contribution of the paper constitutes the introduction of a general hyper-heuristic approach, which is suitable for addressing two rather different timetabling problems in health care. It is applicable without much effort, provided a set of low-level heuristics is available for each problem. We consider the instances of both health care problems for testing the general applicability of the hyper-heuristic approach. Also, improvements to the previous best results for the patient admission scheduling problem are presented. Solutions to the new nurse rostering instances are presented and compared with results obtained by an integer linear programming approach.

The Multi-Mode Resource-Constrained Multi-Project Scheduling Problem

Scheduling projects is a difficult and time consuming process, and has far-reaching implications for any organization’s operations. By generalizing various aspects of project scheduling, decision makers are enabled to capture reality and act accordingly. In the context of the MISTA 2013 conference, the first MISTA challenge, organized by the authors, introduced such a general problem model: the Multi-Mode Resource-Constrained Multi-Project Scheduling Problem (MRCMPSP). The present paper reports on the competition and provides a discussion on its results. Furthermore, it provides an analysis of the submitted algorithms, and a study of their common elements. By making all benchmark datasets and results publicly available, further research on the MRCMPSP is stimulated.

The Red–Blue transportation problem

This paper considers the Red–Blue Transportation Problem (Red–Blue TP), a generalization of the transportation problem where supply nodes are partitioned into two sets and so-called exclusionary constraints are imposed. We encountered a special case of this problem in a hospital context, where patients need to be assigned to rooms. We establish the problem’s complexity, and we compare two integer programming formulations. Furthermore, a maximization variant of Red–Blue TP is presented, for which we propose a constant-factor approximation algorithm. We conclude with a computational study on the performance of the integer programming formulations and the approximation algorithms, by varying the problem size, the partitioning of the supply nodes, and the density of the problem.

A study of decision support models for online patient-to-room assignment planning

The present paper studies patient-to-room assignment planning in a dynamic context. To this end, an extension of the patient assignment (PA) problem formulation is proposed, for which two online ILP-models are developed. The first model targets the optimal assignment for newly arrived patients, whereas the second also considers future, but planned, arrivals. Both models are compared on an existing set of benchmark instances from the PA planning problem, which serves as the basic problem setting. These instances are then extended with additional parameters to study the effect of uncertainty on the patients’ length of stay, as well as the effect of the percentage of emergency patients. The results show that the second model provides better results under all conditions, while still being computationally tractable. Moreover, the results show that pro-actively transferring patients from one room to another is not necessarily beneficial.

A Guide-and-Observe Hyper-Heuristic Approach to the Eternity II Puzzle

The present paper considers the optimisation version of the Eternity II puzzle problem and unsigned edge matching puzzles in general. The goal of this optimisation problem is to maximise the number of matching edges in the puzzle. In 2010, the META Eternity II contest awarded the best performing metaheuristic approach to this hard combinatorial optimisation problem. The winning hyper-heuristic of the contest is subject of this paper. Heuristic design decisions are motivated based on the results of extensive experiments. Furthermore, new results for the Eternity II puzzle problem are presented. The main contribution of this paper is the description of a novel guide-and-observe search mechanism combining a set of objectives. The approach significantly outperforms search methods guided by the default objective only.

Searching for a cycle with maximum coverage in undirected graphs

The present contribution considers the problem of identifying a simple cycle in an undirected graph such that the number of nodes in the cycle or adjacent to it, is maximum. This problem is denoted as the Maximum Covering Cycle Problem and it is shown to be NP-complete. We present an iterative procedure that, although it cannot be shown to be polynomial, yields (in practice) high-quality solutions within reasonable time on graphs of moderate density.

The one-dimensional cutting stock problem with sequence-dependent cut losses

The paper presents a new generalisation of the one-dimensional cutting stock problem (1D-CSP) that considers cut losses that depend on the items’ cutting sequence. It is shown that this generalisation can still be solved approximately by standard 1D-CSP approaches. Furthermore, a pattern-based heuristic (denoted HSD) is presented that specifically considers sequence dependent cut losses (SDCL). A computational study shows that whenever some variability in SDCL occurs consideration of SDCL in the HSD heuristic is beneficial. Finally, two case studies illustrate the relevance of this new generalisation.

Context-Aware and Self-learning Dynamic Transport Scheduling in Hospitals

The increase in available ICT infrastructure in hospitals offers cost reduction opportunities by optimizing various workflows, while maintaining quality of care. In this demonstrator-paper, we present a self-learning dashboard, for monitoring and learning the cause of delays of hospital transports. By identifying these causes, future delays in transport time can be reduced.