EVOLVE International Conference

EVOLVE 2014 July 1-4
Beijing, China

EVOLVE2014 Sessions

Machine Learning Applied to Networks (MLAN)

NEW DEADLINE:  February 28, 2014

Organizers: Slawomir Kuklinski (chair), Emilia Tantar (co-chair)


Nowadays communication networks work by relying on a limited number of communication protocols, namely responsible for data switching or routing. These protocols are typically based on some heuristic assumptions and rarely provide an optimized network behaviour. Moreover, they only provide a limited functionality. As a consequence, additional network operation support is needed, typically related to network configuration, fault management and performance optimization via policy control. Such management operations are in most cases centralized and often performed manually.

The growing number of nodes and users as well as the increase in complexity of all kinds of communication networks call for a new approach in operation and management. On the other hand, some new networking concepts offer new grounds for the automation of most control and management operations. Namely, they allow optimizing the network performance, increase the robustness and automate the time critical management operations. It is worthwhile mentioning the concepts like Software Defined Networking (SDN), 3GPP Self-organizing Networks (SON), or autonomic networking paradigm. A key role of all these concepts is played by algorithms responsible for the behaviour of the network. The complexity and distributed nature of communication networks require sophisticated algorithms that are able to adapt to changes in network topology, faults, dynamic requirements, users’ mobility and their communication demands. In this context, machine learning based algorithms, that are  able to adapt to a changing environment, may play an important role. The aim of this track is to address all potential applications of machine learning techniques in network control and management operations, including the following technologies:

  • 3GPP Self-organizing Networks (SON)
  • Software Defined Networking (SDN)
  • Energy efficient enabled networking
  • IP protocols based networks
  • Mobile networks
  • Autonomic and cognitive network management
  • Wireless networks
  • Sensor networks
  • Networks security

Workshop Chair: Slawomir Kuklinski, Warsaw University of Technology and Orange Labs Poland
Workshop Co-chair: Emilia Tantar, University of Luxembourg

Technical Program Committee:

  1. Prosper Chemouil, Orange Labs, France
  2. Radu State, UL, Luxembourg
  3. Latif Ladid, UL, IPv6 Forum, Luxembourg
  4. Thorsten Ries, UL, Luxembourg
  5. Zwi Altman, Orange, France
  6. Petteri Mannersalo, VTT, Finland
  7. Jarosław Arabas, WUT Poland
  8. Jacek Wytrębowicz, WUT, Poland
  9. Andrzej Duda, INP Grenoble, France
  10. Alex Galis, UCL, United Kingdom
  11. Brian Lee, AIT, Ireland
  12. Tony Jokikyyny, Ericsson Finland
  13. Wendong Wang, BUPT, China
  14. Yuhong Li, BUPT, China
  15. Aki Nakao, University of Tokyo
  16. Ranganai Chaparadza, IPv6 Forum
  17. Tija Ojanpera, VTT, Finland
  18. Vilho Raisanen, NSN, Finland
  19. Minho Jo, South Korea


Sławomir Kukliński

SlawomirKuklinskiDr. Sławomir Kukliński received Ph.D. degree in Telecommunications from Warsaw University of Technology (WUT), Institute of Telecommunications, in 1994 with honors and since then he is Assistant Professor there. He is teaching about mobile and wireless systems. From 2003 he works also for Telekomunikacja Polska R&D Centre (at present it is Orange Labs Poland) as research expert. He has 25 years long experience in telecommunications. At present he is focused on Future Internet and mobile and wireless systems evolution. He led many national research projects as principal investigator. In his career he was involved in many international projects, including EU funded FP6 MIDAS project concerning context aware routing, FP7 project EFIPSANS concerned autonomic management, the ProSense project on sensor networks, and in FP7 project 4WARD in which he was working on new paradigms of network management. At present he is playing a key role in the Celtic COMMUNE Project (2011-2014), which concerns cognitive network management under uncertainty and from 2013 he is the coordinator of 3 years long Polish-Luxembourgish project on Cognitive SDN (CoSDN). Slawomir Kuklinski also participates in ITU-T standardization (Study Group 13). He published many conference and journal papers, was a member of TPC of many conferences, and served as a reviewer to many conferences and journals, including IEEE Communications Magazine.


Practical Aspects of Evolutionary Algorithms

Organizer: Jorn Mehnen

Date: July 3


The Practical Aspects of Evolutionary Computation (PractEC) track at EVOLVE 2014 is
dedicated to the discussion of issues related to real-world application of Evolutionary Computation. This track tries to bridge the gap between fundamental science and industrial applications. It discusses how EC-related technologies are being used to solve real-world problems.

The PractEC track serves several purposes:

- provides a forum in which practitioners of real-world systems can describe their approaches in a non-technical way
- serves as a platform for academics who want to introduce practical solutions and problems to industry
- provides top-level presentations of techniques across industries and organisations
- serves as a place for industrial attendees to introduce their EC approaches
- helps students learn how to get employment in EC

Schedule of PractEC:

July 3

10:00 - 11:00       Ask the Experts
11:15 - 12:15       Applied EC Papers
14:00 - 15:00       Real-world talks
15:00 - 16:00       How to find a Job in Evolutionary Computing

Ask the Experts

This is an excellent opportunity to openly discuss any practical issues with EC in a forum of experts with decades of experience. This is also a great opportunity to get in contact with industry and future employees for social and technical networking.

Ask straight away or listen in … you may just find the answer to your current problem or find the right stimulation for new ideas.

Applied EC papers and Real-world talks

This is a forum to present selected practical papers related to real-world applications of evolutionary computation.

Selected speakers from industry, governmental bodies and academia will introduce their very practical views and results around applied real-world EC projects.

How to find a Job in Evolutionary Computing

Getting a job with training in evolutionary computation can be much easier if you know the things to do and the things not to do in your last year or two of study. In this session you will hear from a panel of experts who have trained students and who
have hired students to carry out real-world optimisation.

Highly recommended if you will be looking for a job in the next few years - or if you are thinking of changing jobs.

Computational Game Theory

Organizers: Rodica Ioana Lung and D. Dumitrescu


Standard Game theory offers mathematical tools for modeling strategic/conflicting agent interactions and proposes characterizations of equilibria (solution concepts) for such situations with multiple applications in economics, computer science, biology, engineering, etc. Computational Game Theory enhances these concepts by offering practical tools for equilibria detection and simulations  and by providing decision makers with functional solutions for these situations. Due to their specific flexibility, adaptability and scalability, artificial intelligence tools have the potential to deal with the complex issues arising in this field.

The aim of the special session in CGT is to bring together researchers working in this field in order to exchange ideas and to promote deeper investigations on game theory approaches by computational intelligence methods.

SS in CGT invites researchers to submit their papers related but not limited to the following topics:

  • Equilibrium Concepts and Detection
  • Evolutionary Game Theory
  • Iterated/repeated Games
  • Dynamic Games
  • Evolutionary algorithms for Game Equilibrium Detection
  • Behavioral Game Theory
  • Algorithmic Game Theory
  • Applications of Game Theory
  • Cooperative Games
  • Network Games and Graph-Theoretic Aspects of Social Networks
  • Solving Complex problems by using Game Theory
  • Decentralized equilibrium detection
  • Equilibrium emergence in multi agent systems
  • Strategic interactions

Robust Optimization

Organizer: Massimiliano Vasile


Uncertainty quantification is a growing field of research, however its integration with computational optimization is still an emerging technology with great potential to address some of the challenges that our world is facing. The fusion of uncertainty quantification and computational optimization, generally referred to as robust optimization or optimization under uncertainty, is expected to have a significant impact on: engineering design and manufacturing (optimal design of product and processes under uncertainty), climate change and control (optimal design of human activities under uncertainty), finance, system biology, drug protocol definition, energy production and distribution, sustainable building design and control, just to mention a few key areas. In an expanding world with limited resources, optimization becomes essential. However, our representation of the world is often incomplete and affected by errors. The inclusion of uncertainties in the optimization process is therefore of primary importance to design solutions that are weakly affected by uncertainties and failure averse.

This session aims at promoting a constructive discussion on open problems, required advances and possible applications of computational optimization and uncertainty quantification with the precise goal to identify new avenues of collaborative research and development.

You are kindly invited to contribute to the session with a talk, and a paper, on the
following subjects:
- Design for reliability
- Optimization Under Uncertainty
- Imprecise Probabilities
- Subjective Probabilities and Elicitation
- Applications of Uncertainty Quantification and Robust Design to Real Engineering Problems
- Uncertainty Quantification and Robust Optimization in Manufacturing Processes
- Climate Uncertainty Quantification and Control
- Robust Design of Drugs and Drug Protocols
- Uncertainty Quantification in Model-Based Design
- Uncertainty Quantification and Data Analysis for Inverse Problems
- Model Reduction
- Uncertainty Quantification and Optimization of Dynamical Systems
- Uncertainty Quantification in Multi-fidelity Approaches

Genetic Programming

Organizers: Leonardo Trujillo, Edgar Galvan-Lopez and Pierrick Legrand


The field of Genetic Programming (GP) studies the development of evolutionary algorithms that synthesize computer programs to automatically solve a specific computational problem. The flexibility and power of GP is by now evident, with the development of successful applications in a variety of domains, from pure mathematics, to signal processing, engineering and software repair. However, important open questions still need to be addressed to overcome some of the limitations of current GP-based methods, which can only be achieved with a deeper understanding of the fundamental dynamics of a GP search and the development of hybrid techniques that incorporate traditional machine learning paradigms. In this special session we invite research contributions that study both experimental and theoretical aspects of GP, as well as innovative applications.

This special session is intended as a platform for studies on GP. Topics include, but are not limited to:

- Problem difficulty
- Novel representations and operators
- Convergence models
- Bloat
- Hybrid methods
- Fitness landscape analysis
- Real-world applications
- Software Engineering and GP

Joomla templates by Joomlashine