Michael Emmerich

Michael Emmerich
University of Jyväskylä | JYU · Department of Mathematical Information Technology

Dr. rer. nat.
Lead AI Scientist, Silo.ai University Researcher, University of Jyväskylä

About

335
Publications
89,909
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
8,102
Citations
Citations since 2017
151 Research Items
5446 Citations
201720182019202020212022202302004006008001,000
201720182019202020212022202302004006008001,000
201720182019202020212022202302004006008001,000
201720182019202020212022202302004006008001,000
Introduction
Lead AI Scientist at Silo AI, Jyväkylä, FI. Jyväskylä University, Finland (Multiobjective Optimization Group) Former Associate Prof. Leiden University Jus Promovendi at Leiden University Interests: Computational Geometry, Multiobjective Optimization, Computational Chemistry and Chemical Engineering, Complex Networks
Additional affiliations
September 2019 - present
University of Jyväskylä
Position
  • Professor
Description
  • Guest Researcher
August 2012 - present
Leiden University
Position
  • Multiobjective Optimization and Decision Making (MSc Course)
June 2010 - January 2011
Universidade do Algarve
Position
  • Set Indicators in Multiobjective Optimization (SIMO)

Publications

Publications (335)
Article
Complexity in solving real-world multicriteria optimization problems often stems from the fact that complex, expensive and/or time-consuming simulation tools or physical experiments are used to evaluate solutions to a problem. In such settings it is common to use efficient computational models, often known as surrogates or meta-models, to approxima...
Conference Paper
Full-text available
The Expected Hypervolume Improvement (EHVI) is a frequently used infill criterion in surrogate-assisted multi-criterion optimization. It needs to be frequently called during the execution of such algorithms. Despite recent advances in improving computational efficiency, its running time for three or more objectives has remained in \(O(n^d)\) for \(...
Conference Paper
Full-text available
Let B be a set of n axis-parallel boxes in d-dimensions such that each box has a corner at the origin and the other corner in the positive quadrant, and let k be a positive integer. We study the problem of selecting k boxes in B that maximize the volume of the union of the selected boxes. The research is motivated by applications in skyline queries...
Article
Full-text available
The management of epidemics received much interest in recent times, due to devastating outbreaks of epidemic diseases such as Ebola and COVID-19. This paper investigates the effect of the structure of the contact network on the dynamics of the epidemic outbreak. In particular we focus on the peak number of critically infected nodes, because this de...
Chapter
In this chapter, we present a Many-Criteria Optimisation and Decision Analysis (MACODA) Ontology and MACODA Knowledge Management Web-Based Platform (named MyCODA, available at http://macoda.club) for the research community. The purpose of this initiative is to allow for the collaborative development of an ontology to represent the MACODA knowledge...
Chapter
Many-objective optimization problems (MaOPs) are problems that feature four or more objectives, criteria or attributes that must be considered simultaneously. MaOPs often arise in real-world situations and the development of algorithms for solving MaOPs has become one of the hot topics in the field of evolutionary multi-criteria optimization (EMO)....
Chapter
Full-text available
Next to the Pareto dominance relation, alternative order relations can be useful in many-objective optimization. In particular, it is interesting to extend the Pareto dominance relation in order to make more pairs comparable and decrease the size of the optimal set (for discrete approximations in continuous optimization or discrete problems), which...
Article
Full-text available
MultiOptForest is an open-source software designed to simplify building and solving multi-objective optimization problems for forest planning. It aims to find the optimal portfolio of management regimes that balance the objectives regarding multiple forest ecosystem services and biodiversity. The software flexibly imports data, allowing for the use...
Article
Full-text available
The European Union (EU) set clear climate change mitigation targets to reach climate neutrality, accounting for forests and their woody biomass resources. We investigated the consequences of increased harvest demands resulting from EU climate targets. We analysed the impacts on national policy objectives for forest ecosystem services and biodiversi...
Chapter
Airline crew pairing optimization problem (CPOP) aims to find a set of flight sequences (crew pairings) that cover all flights in an airline’s highly constrained flight schedule at minimum cost. Since crew cost is second only to the fuel cost, CPOP solutioning is critically important for an airline. However, CPOP is NP-hard, and tackling it is quit...
Chapter
The problem of approximating the Pareto front of a multiobjective optimization problem can be reformulated as the problem of finding a set that maximizes the hypervolume indicator. This paper establishes the analytical expression of the Hessian matrix of the mapping from a (fixed size) collection of n points in the d-dimensional decision space (or...
Chapter
Full-text available
A building spatial design (BSD) determines external and internal walls and ceilings of a building. The design space has a hierarchical structure, in which decisions on the existence or non-existence of spatial components determine the existence of variables related to these spaces, such as sizing and angles. In the optimization of BSDs it is envisi...
Article
The factors determining a drug's success are manifold, making de novo drug design an inherently multi-objective optimisation (MOO) problem. With the advent of machine learning and optimisation methods, the field of multi-objective compound design has seen a rapid increase in developments and applications. Population-based metaheuris-tics and deep r...
Article
Full-text available
Recently, the Hypervolume Newton Method (HVN) has been proposed as a fast and precise indicator-based method for solving unconstrained bi-objective optimization problems with objective functions. The HVN is defined on the space of (vectorized) fixed cardinality sets of decision space vectors for a given multi-objective optimization problem (MOP) an...
Article
Full-text available
In recent years, interactive evolutionary multiobjective optimization methods have been getting more and more attention. In these methods, a decision maker, who is a domain expert, is iteratively involved in the solution process and guides the solution process toward her/his desired region with preference information. However, there have not been m...
Preprint
The problem of approximating the Pareto front of a multiobjective optimization problem can be reformulated as the problem of finding a set that maximizes the hypervolume indicator. This paper establishes the analytical expression of the Hessian matrix of the mapping from a (fixed size) collection of $n$ points in the $d$-dimensional decision space...
Preprint
Recently, the Hypervolume Newton method (HVN) has been proposed as fast and precise indicator-based method for solving unconstrained bi-objective optimization problems with objective functions that are at least twice continuously differentiable. The HVN is defined on the space of (vectorized) fixed cardinality sets of decision space vectors for a g...
Article
Full-text available
Optimization problems with multiple objectives and many input variables inherit challenges from both large-scale optimization and multi-objective optimization. To solve the problems, decomposition and transformation methods are frequently used. In this study, an improved control variable analysis is proposed based on dominance and diversity in Pare...
Article
The landmark achievements of AlphaGo Zero have created great research interest into self-play in reinforcement learning. In self-play, Monte Carlo Tree Search (MCTS) is used to train a deep neural network, which is then used itself in tree searches. The training is governed by many hyper-parameters. There has been surprisingly little research on de...
Preprint
Full-text available
Automated model selection is often proposed to users to choose which machine learning model (or method) to apply to a given regression task. In this paper, we show that combining different regression models can yield better results than selecting a single ('best') regression model, and outline an efficient method that obtains optimally weighted con...
Preprint
Full-text available
This work provides the exact expression of the probability distribution of the hypervolume improvement (HVI) for bi-objective generalization of Bayesian optimization. Here, instead of a single-objective improvement, we consider the improvement of the hypervolume indicator concerning the current best approximation of the Pareto front. Gaussian proce...
Preprint
Full-text available
This paper investigates the effect of the structure of the contact network on the dynamics of the epidemic outbreak. In particular, we focus on the peak number of critically infected nodes (PCIN), determining the maximum workload of intensive healthcare units which should be kept low. As a model and simulation method, we develop a continuous-time M...
Article
The most relevant property that a quality indicator (QI) is expected to have is Pareto compliance, which means that every time an approximation set strictly dominates another in a Pareto sense, the indicator must reflect this. The hypervolume indicator and its variants are the only unary QIs known to be Pareto-compliant but there are many commonly...
Article
Full-text available
We introduce novel concepts to solve multiobjective optimization problems involving (computationally) expensive function evaluations and propose a new interactive method called O-NAUTILUS. It combines ideas of trade-off free search and navigation (where a decision maker sees changes in objective function values in real time) and extends the NAUTILU...
Article
Full-text available
In polypharmacology drugs are required to bind to multiple specific targets, for example to enhance efficacy or to reduce resistance formation. Although deep learning has achieved a breakthrough in de novo design in drug discovery, most of its applications only focus on a single drug target to generate drug-like active molecules. However, in realit...
Article
Multi-objective (MO) optimization, i.e., the simultaneous optimization of multiple conflicting objectives, is gaining more and more attention in various research areas, such as evolutionary computation, machine learning (e.g., (hyper-)parameter optimization), or logistics (e.g., vehicle routing). Many works in this domain mention the structural pro...
Article
A preference based multi-objective evolutionary algorithm is proposed for generating solutions in an automatically detected knee point region. It is named Automatic Preference based DI-MOEA (AP-DI-MOEA) where DI-MOEA stands for Diversity-Indicator based Multi-Objective Evolutionary Algorithm). AP-DI-MOEA has two main characteristics: firstly, it ge...
Article
Vehicle fleets support a diverse array of functions and are increasing rapidly in the world of today. For a vehicle fleet, maintenance plays a critical role. In this article, an evolutionary algorithm is proposed to optimize the vehicle fleet maintenance schedule based on the predicted remaining useful lifetime (RUL) of vehicle components to reduce...
Preprint
p>In polypharmacology, ideal drugs are required to bind to multiple specific targets to enhance efficacy or to reduce resistance formation. Although deep learning has achieved breakthrough in drug discovery, most of its applications only focus on a single drug target to generate drug-like active molecules in spite of the reality that drug molecules...
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...
Article
Full-text available
For almost 20 years, quality indicators (QIs) have promoted the design of new selection mechanisms of multi-objective evolutionary algorithms (MOEAs). Each indicator-based MOEA (IB-MOEA) has specific search preferences related to its baseline QI, producing Pareto front approximations with different properties. In consequence, an IB-MOEA based on a...
Preprint
Full-text available
A preference based multi-objective evolutionary algorithm is proposed for generating solutions in an automatically detected knee point region. It is named Automatic Preference based DI-MOEA (AP-DI-MOEA) where DI-MOEA stands for Diversity-Indicator based Multi-Objective Evolutionary Algorithm). AP-DI-MOEA has two main characteristics: firstly, it ge...
Preprint
The general problem in this paper is vertex (node) subset selection with the goal to contain an infection that spreads in a network. Instead of selecting the single most important node, this paper deals with the problem of selecting multiple nodes for removal. As compared to previous work on multiple-node selection, the trade-off between cost and b...
Book
The 14th International Workshop on Global Optimization was organized by Leiden University (Leiden Centre for Advanced Computer Science and Mathematical Institute) and the International Society of Global Optimization. One of the highlights of this workshop was a particular focus on the topic of multiobjective global optimization. LeGO 2018 is a work...
Book
LEGO Global Optimization Workshop Proceedings, Leiden 2018
Chapter
Full-text available
This work investigates the effect of information exchange in decomposition methods that work with multi-membered populations as sub-problems. As an algorithm framework, we use the Multi-objective Evolutionary Algorithm based on Sub-populations (MOEA/S). This algorithm uses parallel sub-populations that can exchange information via migration and/or...
Chapter
Full-text available
Given a point in m-dimensional objective space, any -ball of a point can be partitioned into the incomparable, the dominated and dominating region. The ratio between the size of the incomparable region, and the dominated (and dominating) region decreases proportionally to , i.e., the volume of the Pareto dominating orthant as compared to all other...
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
For a large-scale airline, the crew operating cost is second only to the fuel cost. This makes the role of crew pairing optimization (CPO) critical for business viability. Here, the aim is to generate a set of flight sequences (crew pairings) that cover a finite set of an airline’s flight schedule, at minimum cost, while satisfying several legality...
Preprint
Full-text available
Morpion Solitaire is a popular single player game, performed with paper and pencil. Due to its large state space (on the order of the game of Go) traditional search algorithms, such as MCTS, have not been able to find good solutions. A later algorithm, Nested Rollout Policy Adaptation, was able to find a new record of 82 steps, albeit with large co...
Preprint
Given a point in m-dimensional objective space, the local environment of a point can be partitioned into the incomparable, the dominated and the dominating region. The ratio between the size of the incomparable region, and the dominated (and dominating) decreases proportionally to $1/2^{m-1}$. Due to this reason, it gets increasingly unlikely that...
Preprint
Full-text available
A customized multi-objective evolutionary algorithm (MOEA) is proposed for the multi-objective flexible job shop scheduling problem (FJSP). It uses smart initialization approaches to enrich the first generated population, and proposes various crossover operators to create a better diversity of offspring. Especially, the MIP-EGO configurator, which...
Preprint
Full-text available
Crew pairing optimization (CPO) is critically important for any airline, since its crew operating costs are second-largest, next to the fuel-cost. CPO aims at generating a set of flight sequences (crew pairings) covering a flight-schedule, at minimum-cost, while satisfying several legality constraints. For large-scale complex flight networks, billi...
Preprint
Full-text available
The landmark achievements of AlphaGo Zero have created great research interest into self-play in reinforcement learning. In self-play, Monte Carlo Tree Search is used to train a deep neural network, that is then used in tree searches. Training itself is governed by many hyperparameters.There has been surprisingly little research on design choices f...
Preprint
Full-text available
Airline scheduling poses some of the most challenging problems in the entire Operations Research (OR) domain. In that, crew scheduling (CS) constitutes one of the most important and challenging planning activities. Notably, the crew operating cost is the second-largest component of an airline's total operating cost (after the fuel cost). Hence, its...
Preprint
Full-text available
Airline crew cost is the second-largest operating cost component and its marginal improvement may translate to millions of dollars annually. Further, it's highly constrained-combinatorial nature brings-in high impact research and commercial value. The airline crew pairing optimization problem (CPOP) is aimed at generating a set of crew pairings, co...
Article
Full-text available
Kriging or Gaussian Process Regression is applied in many fields as a non-linear regression model as well as a surrogate model in the field of evolutionary computation. However, the computational and space complexity of Kriging, that is cubic and quadratic in the number of data points respectively, becomes a major bottleneck with more and more data...
Chapter
Bayesian Global Optimization (BGO) (also referred to as Bayesian Optimization, or Efficient Global Optimization (EGO)), uses statistical models—typically Gaussian process regression to approximate an expensive objective function. Based on this prediction an infill criterion is formulated that takes into account the expected value and variance. BGO...
Book
This two-volume set LNCS 12269 and LNCS 12270 constitutes the refereed proceedings of the 16th International Conference on Parallel Problem Solving from Nature, PPSN 2020, held in Leiden, The Netherlands, in September 2020. The 99 revised full papers were carefully reviewed and selected from 268 submissions. The topics cover classical subjects such...
Book
This two-volume set LNCS 12269 and LNCS 12270 constitutes the refereed proceedings of the 16th International Conference on Parallel Problem Solving from Nature, PPSN 2020, held in Leiden, The Netherlands, in September 2020. The 99 revised full papers were carefully reviewed and selected from 268 submissions. The topics cover classical subjects such...
Conference Paper
Full-text available
Recently, AlphaZero has achieved outstanding performance in playing Go, Chess, and Shogi. Players in AlphaZero consist of a combination of Monte Carlo Tree Search and a deep neural network, that is trained using self-play. The unified deep neural network has a policy-head and a value-head, and during training, the optimizer minimizes the sum of pol...
Article
Full-text available
Abstract Drugs have become an essential part of our lives due to their ability to improve people’s health and quality of life. However, for many diseases, approved drugs are not yet available or existing drugs have undesirable side effects, making the pharmaceutical industry strive to discover new drugs and active compounds. The development of drug...
Article
Full-text available
In the field of multi-objective optimization algorithms, multi-objective Bayesian Global Optimization (MOBGO) is an important branch, in addition to evolutionary multi-objective optimization algorithms. MOBGO utilizes Gaussian Process models learned from previous objective function evaluations to decide the next evaluation site by maximizing or min...
Chapter
Full-text available
After the recent groundbreaking results of AlphaGo and AlphaZero, we have seen strong interests in deep reinforcement learning and artificial general intelligence (AGI) in game playing. However, deep learning is resource-intensive and the theory is not yet well developed. For small games, simple classical table-based Q-learning might still be the a...
Conference Paper
Full-text available
The technique of parallelization is a trend in the field of Bayesian global optimization (BGO) and is important for real-world applications because it can make full use of CPUs and speed up the execution times. This paper proposes a multi-point mechanism of the expected hypervolume improvement (EHVI) for multi-objective BGO (MOBGO) by the utilizati...
Conference Paper
Full-text available
Effective soil-sampling is essential for the construction of prescription maps used in Precision Agriculture for Variable Rate Application of nutrients. In practice, designing a field sampling plan is subject to hard limitations, merely due to the associated expenses, where only a few sample points are taken for evaluation. The accuracy of construc...
Conference Paper
In this paper we propose a tabu search-based memetic algorithm (TSM) for the multi-objective flexible job shop scheduling problem (FJSSP), with the objectives to minimize the makespan, the total workload and the critical workload. The problem is addressed in a Pareto manner, which targets a set of Pareto optimal solutions. The novelty of our method...
Conference Paper
Artificial neural networks typically use backpropagation methods for the optimization of weights. In this paper, we aim at investigating the potential of applying the so-called evolutionary strategies (ESs) on the weight optimization task. Three commonly used ESs are tested on a multilayer feedforward network, trained on the well-known MNIST data s...
Conference Paper
Full-text available
The performance comparison of multi-objective evolutionary algorithms (MOEAs) has been a broadly studied research area. For almost two decades, quality indicators (QIs) have been employed to quantitatively compare the Pareto front approximations produced by MOEAs. QIs are set-functions that assign a real value, depending on specific preferences, to...