Data and results from the paper “Deriving rules of thumb for facility decision making in humanitarian operations” | July 4, 2020

Below, you can find the summary of results of a large experimental study from the paper:

In this paper, we investigate the factors that have an impact on the choice of facility configuration for inventory pre-positioning in preparation for emergencies – a critical decision faced by humanitarian managers. Current research in the field is rich with mathematical models and solution algorithms for the problem of pre-positioning emergency supplies. However, due to a lack of a strong mathematical background and/or computational infrastructure, the decision makers rely on simpler rules of thumb to guide their planning. Some managerial implications have been offered in the literature, but these have been derived from sensitivity analyses focused on a single factor and using a single case study, and as such can be misleading as they ignore important interactions between many instance characteristics.

On the one hand, the outcomes of the study help us identify the most important factors and factor interactions that are further used to yield policy recommendations for facility planning. On the other hand, this study also demonstrates the extent of erroneousness of the guidelines derived from simple analyses, and as such hopefully promotes better experimental designs in the field of humanitarian logistics

Meta-analysis of metaheuristics: A path to generalizable knowledge | November 9, 2019

The findings of this meta-analysis underline the importance of evaluating the contribution of metaheuristic components, and of knowledge over competitive testing. Our goal with the aforementioned paper was to promote meta-analysis as a methodology of obtaining knowledge and understanding of metahueristics frameworks, and we hope to see an increase in its popularity in the domain of operations research.

Research on metaheuristics has focused almost exclusively on (novel) algorithmic development and on competitive testing, both of which have been frequently argued to yield very little generalizable knowledge. One way to obtain problem- and implementation-independent insights about metaheuristics is meta-analysis, a systematic statistical examination that combines the results of several independent studies. Meta-analysis is widely used in several scientific domains, most notably the medical sciences (e.g., to establish the efficacy of a certain treatment), but has not yet been applied in operations research.

In order to demonstrate its potential in learning about algorithms, we carried out a meta-analysis of the adaptive layer in adaptive large neighborhood search (ALNS). Although ALNS has been widely used to solve a broad range of problems, it has not yet been established whether or not adaptiveness actually contributes to the performance of an ALNS algorithm. A total of 134 studies were identified through Google Scholar or personal e-mail correspondence with researchers in the domain, 63 of which fit a set of predefined eligibility criteria. After sending requests for data to the authors of the eligible studies, results for 25 different implementations of ALNS were collected and analysed using a random-effects model.

The detailed comparison of ALNS with the non-adaptive variant per study and per instance, together with the meta-analysis summary results is publicly available on Mendeley. The data allows to replicate the analysis, to evaluate the algorithms using other metrics, to revisit the importance of ALNS adaptive layer if results from more studies become available, or to simply consult the ready-to-use formulas in META-ANALYSIS.ods to carry out a meta-analysis of any research question.

On average, the addition of an adaptive layer in an ALNS algorithm improves the objective function value by 0.14% (95% confidence interval 0.07 to 0.22%). Although the adaptive layer can (and in a limited number of studies does) have an added value, it also adds considerable complexity and can therefore only be recommended in some very specific situations.

The findings of this meta-analysis underline the importance of evaluating the contribution of metaheuristic components, and of knowledge over competitive testing. Our goal with the aforementioned paper was to promote meta-analysis as a methodology of obtaining knowledge and understanding of metahueristics frameworks, and we hope to see an increase in its popularity in the domain of operations research.

Instances and a highly-effective solver for the location routing problem | May 19, 2019

The location routing problem (LRP) unites two important challenges in the design of distribution systems. On the one hand, the delivery of goods to customers needs to be planned as effectively as possible, and on the other hand, the location of depots from where these deliveries are executed has to be determined carefully. The LRP is of particular importance when it comes to designing complex distribution systems in urban areas to answer questions like: “In which part of the city should micro-hubs be positioned” or “Where should depots be opened to optimize the distribution of parcels to customers?”

We investigated which properties optimally-located depots should have. Our findings led to the development of a highly-effective heuristic to solve a wide range of problems in which location decisions are paired with routing decisions: How to find the best depots among 1000 potential options? How to optimize supply chains with multiple echelons? If we made you curious, have a look at our working paper.

The first version of the corresponding Java program with interface can be downloaded here.

Ready to improve supply chains and cities yourself? Work on challenging state-of-the-art instances: classical instances,  new instances,  large-scale instances, and 2-echelon instances.

Case studies and random instances for the problem of pre-positioning emergency supplies | October 11, 2018

Below, you can find the instances introduced in the paper:

By carefully manipulating some of the instance parameters, we generated 30 case studies that were inspired by 4 instances collected from the literature, that focus on disasters of different type and scale that occurred in different parts of the world. To further diversify the problem library, we developed a tool to algorithmically generate arbitrarily many diverse random instances of any size. The folder instances.zip contains the 30 case studies and 10 random instances mentioned in the paper, but the code for the random instance generator is also made available, so that other random instances can be constructed, and/or the definition of the instances can be modified. For more information on how to construct other random instances, please refer to the READ_ME.txt file.

The instances can be used to support further research on the pre-positioning and related problems.

RoutingSolver – Obtain high-quality solutions in a short time | October 2, 2018

RoutingSolver is an efficient solver for routing problems. Based entirely on local search and problem-knowledge, it computes high-quality solutions  in a few seconds or minutes, while competing with the best algorithms in the field. It can also be applied to obtain routing solutions for very large instances with tens of thousands of customers, and solve routing variants of the Vehicle Routing Problem with mutiple depots, multiple trips and location decisions.

The corresponding Java program with interface can be downloaded here.

Kenneth Sörensen over de algoritmes achter het deelfietsensysteem.

ELEVATOR PITCH

Hoe leert een computer denken als Beethoven?

Universiteit van Vlaanderen

Kraken sudoku-makers weldra je bankgegevens?

Universiteit van Vlaanderen