Constraint multi-objective optimization with a limited budget for function evaluations is challenging. This thesis tackles this problem by proposing new optimization algorithms. These algorithms are applied on holistic ship design problems. This helps naval architects balance objectives like cost, efficiency,…

A common method to solve expensive function evaluation problem is using Bayesian Global Optimization, instead of Evolutionary Algorithms.

A modern ship design process is subject to a wide variety of constraints such as safety constraints, regulations, and physical constraints.

This thesis seeks to combine two different research topics; Multi-Objective Optimization and Complex Network Analysis.

Promotores: Prof.dr. T.H.W. Bäck, Prof.dr. M.R.V. Chaudron, Co-Promotor: M.T.M. Emmerich

Promotores: Prof.dr. T.H.W. Bäck, Prof.dr. M. Bongards (Cologne University)

Optimization tasks in practice have multifaceted challenges as they are often black box, subject to multiple equality and inequality constraints and expensive to evaluate.

It is a common technique in global optimization with expensive black-box functions to learn a surrogate-model of the response function from past evaluations and use it to decide on the location of future evaluations.

This thesis covers several aspects of quantum algorithms for near-term quantum computers and its applications to quantum chemistry and material science.

Het onderzoek in het natural computing cluster omvat theoretische fundamenten, de ontwikkeling van nieuwe algoritmen en interdisciplinaire toepassingen van natuurlijke computermethoden. De drijvende kracht achter hun onderzoek is de missie om het begrip van natuurlijke systemen als rekenmodellen te…

In this thesis, topics relating to the optimization of high-throughput pipelines used for imaging are discussed. In particular, different levels of implementation, i.e., conceptual, software, and hardware, are discussed and the thesis outlines how advances on each level need to be made to make gains…

Continuous optimization is never easy: the exact solution is always a luxury demand and the theory of it is not always analytical and elegant.

Industrial manufacturing processes, such as the production of steel or the stamping of car body parts, are complex semi-batch processes with many process steps, machine parameters and quality indicators.

Multi-objective evolutionary computation aims to find high quality (Pareto optimal) solutions that represent the trade-off between multiple objectives.

This thesis involves three topics: benchmarking discrete optimization algorithms, empirical analyses of evolutionary computation, and automatic algorithm configuration.

Moderne wetenschap en technologie, van klimaatmodellen tot kunstmatige intelligentie, draaien op extreem krachtige computers.

Multi-criteria optimalisatie is een effectieve techniek voor het vinden van optimale oplossingen die een afweging bieden tussen verschillende, tegenstrijdige criteria. Het heeft zijn toepassing gevonden in de wereld om ons heen omdat bij het oplossen van praktische, re¨ele wereld problemen men gewoonlijk…

Promotor: Prof.dr. H.A.G. Wijshoff

Uncertainty and noise are frequently-encountered obstacles in real-world applications of numerical optimization. The practice of optimization that deals with uncertainties and noise is commonly referred to as robust optimization.

Optimally solving real-world expensive Black-Box Optimization (BBO) problems w.r.t. real-world constraints, such as wall-clock time and computational costs, is extremely difficult and tedious.

For optimization problems, it is often unclear how to choose the most appropriate optimization algorithm. As such, rigorous benchmarking practices are critical to ensure we can gain as much insight into the strengths and weaknesses of these types of algorithms.

In dit proefschrift onderzochten we hoe we kankerbehandelingen slimmer en veiliger kunnen doseren. Veel middelen worden nog volgens een “one size fits all”-schema gegeven, vaak gebaseerd op de maximaal verdraagbare dosis. Dat kan leiden tot onnodig veel bijwerkingen, behandelingonderbrekingen en hoge…

Carbon dioxide capture and utilization technologies are necessary to create a truly circular economy. The electrochemical reduction of carbon dioxide to formate is an appealing carbon utilization method as it can be performed at room temperature and pressure, it only requires two electrons, and it has…

The increasing competition in the automotive industry requires the tailored, swift development of technologically sophisticated vehicles. Therefore, the computationally expensive state-of-the-art simulation technologies are combined with optimization algorithms. An example of a real-world optimization…

In design optimization problems, engineers typically handcraft design representations based on personal expertise, which leaves a fingerprint of the user experience in the optimization data. Thus, learning this notion of experience as transferrable design features has potential to improve the performance…

Over several decades, a variety of computational methods for drug discovery have been proposed and applied in practice. With the accumulation of data and the development of machine learning methods, computational drug design methods have gradually shifted to a new paradigm, i.e. deep learning methods…

Quantum annealing belongs to a family of quantum optimization algorithms designed to solve combinatorial optimization problems using programmable quantum hardware. In this thesis, various methods are developed and tested to understand how to formulate combinatorial optimization problems for quantum…

This thesis focuses on addressing four research problems in designing embedded streaming systems.

De Duitse hoogleraar Information Systems and Statistics Heike Trautmann aanvaardde deze maand de Pascalleerstoel bij LIACS, het informatica-instituut aan de Universiteit Leiden. Trautmanns belangrijkste vakgebieden zijn evolutionary multiobjective optimisation en data science, waar ook LIACS sterk in…

Promotor: Prof.dr.ir. E. Deprettere

Automated Machine Learning (AutoML) is widely used to automatically build a suitable practical Machine Learning (ML) model for an arbitrary real-world problem, reducing the effort of practitioners in the ML development cycle for real-world applications. Optimization is a key part of a typical AutoML…

In this dissertation several settings in the Online Learning framework are studied. The first chapter serves as an introduction to the relevant settings in Online Learning and in the subsequent chapters new results and insights are given for both full-information and bandit information settings.

Machine Learning is becoming a more and more substantial technology for industry.

Promotor: Prof.dr. A. P. IJzerman, Co-Promotor: L.H. Heitman

In this work we describe three methods to improve the performance of Quantum Field Theory calculations.

The next generation of high-contrast imaging instruments on space-based observatories requires sophisticated wavefront sensing and control in addition to a high-performance coronagraph.

Conventional food supplies procured from rural areas are increasingly vulnerable to disruptions caused by extreme weather, geopolitical instability, and logistical bottlenecks.

Promotor: G.P. van Wezel, Co-Promotor: E. Vijgenboom

On a structural level, the properties featured by a majority of mechanical metamaterials can be ascribed to the finite number of soft internal degrees-of freedom allowing for low-energy deformations.

Promotors: Prof.dr. T.H.W. Bäck, Prof.dr. B. Sendhoff (Technische Universität Darmstadt)

This thesis investigates how quantum, quantum-inspired, and hybrid quantum-classical computation can enhance key points of the automated machine learning (AutoML) pipeline under the constraints of noisy intermediate-scale quantum (NISQ) devices.

Inverse problems are problems where we want to estimate the values of certain parameters of a system given observations of the system. Such problems occur in several areas of science and engineering. Inverse problems are often ill-posed, which means that the observations of the system do not uniquely…

In deep reinforcement learning, searching and learning techniques are two important components. They can be used independently and in combination to deal with different problems in AI, and have achieved impressive results in game playing and robotics. These results have inspired research into artificial…

This thesis aims to develop model-based approaches to optimize drug dosing in patients with overweight and (morbid) obesity.

This thesis introduces the concept of