A5064553018- Reinforcement Learning in Robotics
- Evolutionary Algorithms and Applications
- Game Theory and Applications
- Advanced Multi-Objective Optimization Algorithms
- Optimization and Search Problems
- Metaheuristic Optimization Algorithms Research
- Auction Theory and Applications
- Advanced Bandit Algorithms Research
- Smart Grid Energy Management
- Energy Efficient Wireless Sensor Networks
- Advanced Optical Network Technologies
- Scheduling and Optimization Algorithms
- Optical Network Technologies
- Fuzzy Logic and Control Systems
- Multi-Agent Systems and Negotiation
- Bioinformatics and Genomic Networks
- Advanced Control Systems Optimization
- Gene expression and cancer classification
- Logic, Reasoning, and Knowledge
- Advanced Photonic Communication Systems
- Adaptive Dynamic Programming Control
- Rough Sets and Fuzzy Logic
- Energy Load and Power Forecasting
- Machine Fault Diagnosis Techniques
- Distributed Control Multi-Agent Systems
Vrije Universiteit Brussel
2015-2024
Université Libre de Bruxelles
2011-2024
Universitat de Miguel Hernández d'Elx
2018
Universidad Politécnica de Madrid
2018
École Polytechnique Fédérale de Lausanne
2018
Flanders Make (Belgium)
2018
Universidad Carlos III de Madrid
2018
CMCL Innovations (United Kingdom)
2006-2018
University of Alicante
2018
KU Leuven
2016
A plenitude of feature selection (FS) methods is available in the literature, most them rising as a need to analyze data very high dimension, usually hundreds or thousands variables. Such sets are now various application areas like combinatorial chemistry, text mining, multivariate imaging, bioinformatics. As general accepted rule, these grouped filters, wrappers, and embedded methods. More recently, new group has been added framework FS: ensemble techniques. The focus this survey on filter...
Genomic data integration is a key goal to be achieved towards large-scale genomic analysis. This process very challenging due the diverse sources of information resulting from genomics experiments. In this work, we review methods designed combine recorded microarray gene expression (MAGE) It has been acknowledged that main source variation between different MAGE datasets so-called ‘batch effects’. The reviewed here perform by removing (or more precisely attempting remove) unwanted associated...
In this paper we introduce a new decentralized digital currency, called NRGcoin. Prosumers in the smart grid trade locally produced renewable energy using NRGcoins, value of which is determined on an open currency exchange market. Similar to Bitcoins, offers numerous advantages over fiat but unlike Bitcoins it generated by injecting into grid, rather than spending computational power. addition, propose novel trading paradigm for buying and selling green grid. Our mechanism achieves demand...
Many real-world problems involve the optimization of multiple, possibly conflicting objectives. Multi-objective reinforcement learning (MORL) is a generalization standard where scalar reward signal extended to multiple feedback signals, in essence, one for each objective. MORL process policies that optimize criteria simultaneously. In this paper, we present novel temporal difference algorithm integrates Pareto dominance relation into approach. This multi-policy learns set dominating single...
Real-world decision-making tasks are generally complex, requiring trade-offs between multiple, often conflicting, objectives. Despite this, the majority of research in reinforcement learning and decision-theoretic planning either assumes only a single objective, or that multiple objectives can be adequately handled via simple linear combination. Such approaches may oversimplify underlying problem hence produce suboptimal results. This paper serves as guide to application multi-objective...
Abstract This work provides a data-oriented overview of the rapidly growing research field covering machine learning (ML) applied to predicting electrochemical corrosion. Our main aim was determine which ML models have been and how well they performed depending on corrosion topic considered. From an extensive review articles presenting comparable performance metrics, ‘Machine for database’ created, guiding experts model developers in their applications Potential gaps recommendations are...
With an abundant amount of microarray gene expression data sets available through public repositories, new possibilities lie in combining multiple existing sets. In this context, analysis itself is no longer the problem, but retrieving and consistently integrating all before delivering it to wide variety tools becomes bottleneck. We present newly released inSilicoMerging R/Bioconductor package which, together with earlier inSilicoDb package, allows consistent retrieval, integration publicly...
In multi-objective problems, it is key to find compromising solutions that balance different objectives. The linear scalarization function often utilized translate the nature of a problem into standard, single-objective problem. Generally, noted such as combination can only in convex areas Pareto front, therefore making method inapplicable situations where shape front not known beforehand, case. We propose non-linear function, called Chebyshev basis for action selection strategies...
We propose an algorithmic framework for multi-objective multi-armed bandits with multiple rewards. Different partial order relationships from optimization can be considered a set of reward vectors, such as scalarization functions and Pareto search. A function transforms the environment into single objective are popular choice in reinforcement learning. Scalarization techniques straightforwardly implemented current bandit framework, but efficiency these algorithms depends very much on their...
In recent years, an increasing number of outbreaks Dengue, Chikungunya and Zika viruses have been reported in Asia the Americas. Monitoring virus genotype diversity is crucial to understand emergence spread outbreaks, both aspects that are vital develop effective prevention treatment strategies. Hence, we developed efficient method classify sequences with respect their species sub-species (i.e. serotype and/or genotype). This tool provides easy-to-use software implementation this new was...
Notwithstanding important advances in the context of single-variant pathogenicity identification, novel breakthroughs discerning origins many rare diseases require methods able to identify more complex genetic models. We present here Variant Combinations Pathogenicity Predictor (VarCoPP), a machine-learning approach that identifies pathogenic variant combinations gene pairs (called digenic or bilocus combinations). show results produced by this method are highly accurate and precise, an...
Effectively incorporating external advice is an important problem in reinforcement learning, especially as it moves into the real world. Potential-based reward shaping a way to provide agent with specific form of additional reward, guarantee policy invariance. In this work we give novel incorporate arbitrary function same guarantee, by implicitly translating dynamic potentials, which are maintained auxiliary value learnt at time. We show that provided captures input expectation, and...
Throughout the last decade, a whole new generation of powered transtibial prostheses and exoskeletons has been developed. However, these technologies are limited by gait phase detection which controls wearable device as function activities wearer. Consequently, is considered to be great importance, achieving high accuracy will produce more precise, stable, safe rehabilitation device. In this paper, we propose novel percent algorithm that can predict full cycle discretised within 1% interval....
A tremendous amount of DNA sequencing data is being produced around the world with ambition to capture in more detail mechanisms underlying human diseases. While numerous bioinformatics tools exist that allow discovery causal variants Mendelian diseases, little no support provided do same for variant combinations, an essential task causes oligogenic ORVAL (the Oligogenic Resource Variant AnaLysis), which presented here, provides answer this problem by focusing on generating networks...
The 2021 edition of AAMAS, the International Conference on Autonomous Agents and Multiagent Systems, took place from 3rd to 7th May (aamas2021.soton.ac.uk). This year it was organized in form a virtual event attracted over 1,000 registered participants. As every year, conference featured an exciting programme contributed talks, keynotes addresses, tutorials, affiliated workshops, doctoral consortium, more.
In this paper we survey the basics of reinforcement learning and (evolutionary) game theory, applied to field multi-agent systems. This contains three parts. We start with an overview on fundamentals learning. Next summarize most important aspects evolutionary theory. Finally, discuss state-of-the-art mathematical connection
Learning automata (LA) were recently shown to be valuable tools for designing multiagent reinforcement learning algorithms. One of the principal contributions LA theory is that a set decentralized independent able control finite Markov chain with unknown transition probabilities and rewards. In this paper, we propose extend algorithm games--a straightforward extension single-agent decision problems distributed problems. We show under same ergodic assumptions original theorem, extended will...
An increasing amount of microarray gene expression data sets is available through public repositories. Their huge potential in making new findings yet to be unlocked by them for large-scale analysis. In order do so it essential that independent studies designed similar biological problems can integrated, insights obtained. These would remain undiscovered when analyzing the individual because well known small number samples used per experiment a bottleneck genomic By statistical power...