Koen van der Blom

Koen van der Blom
Sorbonne Université | UPMC · Laboratoire d'informatique de Paris 6 (LIP6)

PhD

About

39
Publications
4,112
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
187
Citations
Introduction
Koen van der Blom currently works at LIP6, Sorbonne Université. Koen does research on evolutionary computation and meta-algorithms. He is also interested in multi-objective optimisation, as well as swarm intelligence.
Additional affiliations
March 2019 - August 2022
Leiden University
Position
  • PostDoc Position
March 2015 - March 2019
Leiden University
Position
  • PhD Student
Education
September 2012 - November 2014
Leiden University
Field of study
  • Computer Science
September 2008 - June 2012

Publications

Publications (39)
Presentation
Full-text available
The primary aim when designing and evaluating algorithms to solve computational problems is to find high-quality (or valid) solutions to given problem instances, and to do so as efficiently as possible. The performance of most cutting edge algorithms for challenging problems is significantly influenced by the choice of their components and paramete...
Poster
Full-text available
Many fields of computational science advance through improvements in the algorithms used for solving key problems. These advancements are often facilitated by benchmarks and competitions that enable performance comparisons and rankings of solvers. Simultaneously, meta-algorithmic techniques, such as automated algorithm selection and configuration,...
Conference Paper
Full-text available
Machine learning (ML) has become essential to a vast range of applications, while ML experts are in short supply. To alleviate this problem, AutoML aims to make ML easier and more efficient to use. Even so, it is not clear to which extent AutoML techniques are actually adopted in an engineering context, nor what facilitates or inhibits adoption. To...
Poster
Full-text available
Machine learning (ML) has become essential to a vast range of applications, while ML experts are in short supply. To alleviate this problem, AutoML aims to make ML easier and more efficient to use. Even so, it is not clear to which extent AutoML techniques are actually adopted in an engineering context, nor what facilitates or inhibits adoption. To...
Article
Full-text available
Three methods for early-stage building spatial design optimization are presented, demonstrated, and compared for their qualities and limitations. The first, an evolutionary algorithm, can find well-distributed approximations of the Pareto front, but it uses many design evaluations and it can only explore a limited part of the entire design search s...
Preprint
Full-text available
Following the recent surge in adoption of machine learning (ML), the negative impact that improper use of ML can have on users and society is now also widely recognised. To address this issue, policy makers and other stakeholders, such as the European Commission or NIST, have proposed high-level guidelines aiming to promote trustworthy ML (i.e., la...
Preprint
Full-text available
Optimisation algorithms are commonly compared on benchmarks to get insight into performance differences. However, it is not clear how closely benchmarks match the properties of real-world problems because these properties are largely unknown. This work investigates the properties of real-world problems through a questionnaire to enable the design o...
Conference Paper
Full-text available
Background. The increasing reliance on applications with machine learning (ML) components calls for mature engineering techniques that ensure these are built in a robust and future-proof manner. Aim. We aim to empirically determine the state of the art in how teams develop, deploy and maintain software with ML components. Method. We mined both ac...
Article
Full-text available
Two new methods to generate structural system layouts for conceptual building spatial designs are presented. The first method, the design response grammar, uses design rules—configurable by parameters—to develop a structural system layout step by step as a function of a building spatial design's geometry and preliminary assessments of the structura...
Preprint
Full-text available
The increasing reliance on applications with machine learning (ML) components calls for mature engineering techniques that ensure these are built in a robust and future-proof manner. We aim to empirically determine the state of the art in how teams develop, deploy and maintain software with ML components. We mined both academic and grey literature...
Conference Paper
Full-text available
Benchmarks are a useful tool for empirical performance comparisons. However, one of the main shortcomings of existing benchmarks is that it remains largely unclear how they relate to real-world problems. What does an algorithm’s performance on a benchmark say about its potential on a specific real-world problem? This work aims to identify propertie...
Poster
Full-text available
Benchmarks are a useful tool for empirical performance comparisons. However, one of the main shortcomings of existing benchmarks is that it remains largely unclear how they relate to real-world problems. What does an algorithm’s performance on a benchmark say about its potential on a specific real-world problem? This work aims to identify propertie...
Preprint
Full-text available
Benchmarks are a useful tool for empirical performance comparisons. However, one of the main shortcomings of existing benchmarks is that it remains largely unclear how they relate to real-world problems. What does an algorithm's performance on a benchmark say about its potential on a specific real-world problem? This work aims to identify propertie...
Thesis
Full-text available
Multi-objective evolutionary computation aims to find high quality (Pareto optimal) solutions that represent the trade-off between multiple objectives. Within this field there are a number of key challenges. Among others, this includes constraint handling and the exploration of mixed-integer search spaces. This thesis investigates how these challen...
Poster
Full-text available
Sparkle aims to provide a platform for solving computationally challenging problems such as SAT and AI planning. The key idea here is to allow algorithm developers to focus on improving the state of the art, while Sparkle handles the meta-algortihmics. So far Sparkle has handled algorithm selection and scheduling (AutoFolio), and has been applied i...
Presentation
Full-text available
Sparkle aims to provide a platform for solving computationally challenging problems such as SAT and AI planning. The key idea here is to allow algorithm developers to focus on improving the state of the art, while Sparkle handles the meta-algortihmics. So far Sparkle has handled algorithm selection and scheduling (AutoFolio), and has been applied i...
Conference Paper
Building spatial design is in practice a co-evolutionary design process. To optimize a building spatial design, an evolutionary algorithm can be used, but the search space is large and complex. In simulations of co-evolutionary design processes the size and complexity are not hindering the search. Such simulations are in the presented work proposed...
Presentation
Full-text available
Presentation of the paper "Analysing Optimisation Data for Multicriteria Building Spatial Design"
Chapter
Full-text available
In traditional, single objective, optimisation local optima may be found by gradient search. With the recently introduced hypervolume indicator (HVI) gradient search, this is now also possible for multi-objective optimisation, by steering the whole Pareto front approximation (PFA) in the direction of maximal improvement. However, so far it has only...
Chapter
Full-text available
This paper presents a framework in which a building spatial design optimisation toolbox and a building information modelling environment are coupled. The coupling is used in a case study to investigate the possible challenges that hamper the interaction between a designer and an optimisation method within a BIM environment. The following challenges...
Chapter
Full-text available
Domain experts can benefit from optimisation simply by getting better solutions, or by obtaining knowledge about possible trade-offs from a Pareto front. However, just providing a better solution based on objective function values is often not sufficient. It is desirable for domain experts to understand design principles that lead to a better solut...
Conference Paper
Full-text available
Many problems are of a mixed integer nature, rather than being restricted to a single variable type. Although mixed integer algorithms exist for the single-objective case, work on the multi-objective case remains limited. Evolution strategies are stochastic optimisation algorithms that feature step size adaptation mechanisms and are typically used...
Conference Paper
Full-text available
Bayesian Global Optimization (BGO) is a very efficient technique to optimize expensive evaluation problems. However, the application domain is limited to continuous search spaces when using a BGO algorithm. To solve mixed integer problems with a BGO algorithm, this paper adapts the heterogeneous distance function to construct the Kriging models and...
Presentation
Full-text available
Presentation on the paper "Towards Multi-objective Mixed Integer Evolution Strategies"
Presentation
Full-text available
Presentation on the paper ”A new foraging-based algorithm for online scheduling”
Conference Paper
Full-text available
While much work exists on scheduling, literature in the subfield of online scheduling remains sparse. As with many problems, online scheduling has parallels with natural phenomena. Specifically, online scheduling can be seen in the division of labour among colony insects, such as ants. Although multiple different biological models exist for divisio...
Article
Full-text available
Multi-disciplinary optimisation of building spatial designs is characterised by large solution spaces. Here two approaches are introduced, one being super-structured and the other super-structure free. Both are different in nature and perform differently for large solution spaces and each requires its own representation of a building spatial design...
Presentation
Full-text available
Presentation on the abstract ”Hypervolume Gradient Ascent for Memetic Building Spatial Design Optimisation”
Conference Paper
Full-text available
This paper proposes a new method for multi-disciplinary optimisation of building spatial designs in the preliminary design stage. First it discusses two recently developed building spatial design optimisation methods, one using a super-structured approach and the other applying a super-structure free approach. Subsequently, a combination of the two...
Presentation
Full-text available
Presentation no the paper ”Configuring Advanced Evolutionary Algorithms for Multicriteria Building Spatial Design Optimisation”
Conference Paper
Full-text available
In this paper solution approaches for solving the building spatial design optimisation problem for structural and energy performance are advanced on multiple fronts. A new initialisation operator is introduced to generate an unbiased initial population for a tailored version of SMS-EMOA with problem specific operators. Improvements to the mutation...
Poster
Full-text available
Poster presentation on the paper ”Multicriteria Building Spatial Design with Mixed Integer Evolutionary Algorithms”
Conference Paper
Full-text available
This paper proposes a first step towards multidisciplinary design of building spatial designs. Two criteria, total surface area (i.e. energy performance) and compliance (i.e. structural performance), are combined in a multicriteria optimisation framework. A new way of representing building spatial designs in a mixed integer parameter space is used...
Presentation
Full-text available
Presentation on the paper ”A Super-Structure Based Optimisation Approach for Building Spatial Designs”
Conference Paper
Full-text available
Building design can be supported effectively by computer-aided design exploration. This paper investigates optimisation based on a mixed-integer superstructure representation of the search space of building spatial designs. It can take into account parametric as well as topo-logical variations. In the suggested superstructure – the so-called superc...
Thesis
Full-text available
Scheduling in the painting facilities of automotive manufacturing plants is a complex process. The paint loss caused by machines continuously switching colours leads to significant costs. Optimising these schedules while assigning new jobs in real time and dealing with machine break downs is therefore an important issue in the truck painting proble...

Network

Cited By

Projects

Projects (4)
Project
https://se-ml.github.io/
Project
Our goal is to clarify the characteristics of real-world optimization problems and generate a new benchmark test suite based on the questionnaire (https://tinyurl.com/opt-survey) results. We believe such a test suite helps efficiently developing new optimization methods for real-world optimization problems. Results will be made publicly available at https://sites.google.com/view/macoda-rwp