EC-DIGEST-L Archives

March 2021


Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Eric 'Siggy' Scott <[log in to unmask]>
Reply To:
Evolutionary Computation Digest <[log in to unmask]>
Thu, 4 Mar 2021 18:11:52 -0500
text/plain (386 lines)
Evolutionary Computation Digest — Monday, 01 March 2021, Volume
35: Issue 3

SUBMISSION ADDRESS:    [log in to unmask]

(UN)SUBSCRIPTION INSTRUCTIONS:  at the bottom of this email

Today's Topics:

- 05 March: Phd offer: Integration of Machine Learning into the
Resolution of MO-VRPTWs with Applications in Hospital Environment

 CFPs (with submission deadline)
 - 12 April: Evolutionary Algorithms for High Performance Computing
(GECCO Workshop)
 - 12 April: GECCO Competition on Optimal Camera Placement (OCP) and
Unicost Set Covering (USCP)
 - 12 April: Workshop on Understanding Reproducibility in Evolutionary
Computation (GECCO)
 - 30 April: Intl. Conf. on Metaheuristics and Nature Inspired
Computing (META'2021)


April 7–9, 2021, Seville, Spain: EvoStar 2021

July 10–14, 2021, Lille, France: GECCO 2021

Sender: Laetitia Jourdan <[log in to unmask]>
Subject: Phd offer: Integration of Machine Learning into the
Resolution of MO-VRPTWs with Applications in Hospital Environment


**** Context
ORKAD (Operations Research, Knowledge And Data) is a research team
within the OPTIMA thematic group of the CRIStAL research center
(Centre de Recherche en Informatique, Signal et Automatique de Lille)
(UMR CNRS 9189) of the University of Lille (France). The main
objective of the ORKAD team is to simultaneously exploit combinatorial
optimization and data mining in order to solve optimization problems.
Despite the two scientific domains having evolved more or less
independently from each other, the synergy between combinatorial
optimization and data mining offers the opportunity of improving the
performance of optimization methods with help data mining and, on the
other hand, to solve data mining problems more efficiently with the
help of operations research methods [Dhaenens-Jourdan2016]. Our
approaches are mainly based on mono- and multi-objective combinatorial

INOCS (INtegrated Optimization problems with Complex Structure) is an
INRIA’s research team part of the OPTIMA group of CRIStAL in
Lille(France). The INOCS team aims to develop new models, algorithmic
techniques and implementations for problems with complex structure
according to three types of optimization paradigms: mathematical
optimization, bilevel optimization and robust/stochastic optimization.

This thesis is a collaboration between the two teams, ORKAD and INOCS.
The objective of the thesis is to investigate the use of machine
learning techniques to solve multi-objective combinatorial
optimization problems and in particular Multi-Objective Vehicle
Routing Problems with Time Windows.

**** Objective
This thesis aims to investigate the use of machine learning to solve
Multi-Objective Combinatorial Optimization Problems and in particular
the MO-VRPTW [Jozefowiez2008]. A number of studies have emerged in
recent years to integrate learning techniques in optimization
algorithms for routing problems.  These works have shown that
discovering the structural properties of high-quality solutions, can
strongly affect and enhance the performance of heuristic algorithms
for routing problems [Arnold2019c]. It is also well known that
high-quality solutions of a vehicle routing problem are highly similar
to optimal solutions, that is, are structurally close to the global
optima. Generally, a single objective that usually models an
economical aspect is taken into account in the modelisation of the
problem. This may be acceptable when merchandise is transported, but
may not be the case when it comes to the transport of people. In the
latter case, it is essential to take into account the comfort of the
passengers, while assuring to obtain low cost routes to enhance the
performance of the service company.

This leads to a compromise where both aspects are taken into account,
i.e., multi-objective optimization.
First, the transposition to the multi-objective case  of the method
developed in [Arnold2019a] will be studied and experiments will be
conducted on literature benchmarks. Then, the aim will be to
investigate and understand how the properties of high-quality
solutions with respect to optimal solutions adapt to the
multi-objective context, where, usually, the optimization of one
objective results in the detriment of the other(s). This will lead to
the development of novel ways to integrate machine learning in both
exact algorithms and metaheuristics.

***** Application

The applicants should send in one pdf file to
[log in to unmask], [log in to unmask],
[log in to unmask] before the deadline (March 5th):
- CV
- Academic records + Diploma
- Motivation letter
- Reference letter
- if existing publications

The selected candidates will be contacted for interview then invited
to submit their applications to the second stage for end of March.

Pr. Laetitia Jourdan
Team Leader ORKAD
CRIStAL/Université de Lille/CNRS
Cité Scientifique
59655 Villeneuve d’Ascq Cedex
Mail: [log in to unmask]
Phone FST/CRIStAL: +33 (0)3 28 77 85 19

Sender: EC-Digest Editor
Subject: GECCO’21 Evolutionary Algorithm in High Performance Computing
Workshop Call for Papers

Dear colleagues,

We would like to notify you about the upcoming ACM Evolutionary
Algorithms in High Performance Computing Workshop at GECCO 2021, which
is to be held virtually at Lille, France, from July 10-14, 2021.
Please be aware that GECCO will be entirely virtual because of
COVID-19, and so it will be necessary for a pre-recorded presentation
to be provided for accepted papers.

Scope and Topics of Interest

The wide variety of parallel and distributed versions of EAs make them
an ideal candidate for use with HPC systems. Consequently, EAs have
gathered considerable attention for their ability to accelerate
finding solutions for a variety of computationally expensive problem
domains, including reinforcement learning, neural architecture search,
and model calibration for complex simulations. However, use of HPC
resources adds an implicit secondary objective of ensuring those
resources are used efficiently. This means that practitioners have to
make decisions regarding evolutionary algorithms tailored for maximum
HPC resource use, as well as associated software and hardware support.
New EA-oriented HPC benchmarks might also be needed to guide
practitioners in making those decisions.

We are looking for papers on the following sub-topics to facilitate discussion:

 - algorithmic — what novel EA variants best exploit HPC resources?
 - benchmarks — are there HPC specific measures for EA performance?
 - hardware — can we improve use of HPC hardware, such as GPUs?
 - software — what EA software, or software development practices,
best leverage HPC capabilities?

More information can be found at our website, .

Submission and Important Dates

 - Paper submissions open: February 11, 2021
 - Paper submission deadline: April 12, 2021
 - Notification of acceptance: April 26, 2021
 - Camera ready submission: May 3, 2021
 - Author registration deadline: May 3, 2021
 - Conference date: July 10-14, 2021

This email sent on behalf of the EAHPC organizers, and we apologize
for any multiple postings. Any questions concerning submissions can be
addressed to [log in to unmask]

Sender: Julien Lepagnot <[log in to unmask]>
Subject: GECCO 2021 Competition on OCP and USCP - Call for
participation [kind reminder]

Dear colleagues,

This is just a gentle reminder to not forget to register to the GECCO
2021 Competition on OCP and USCP.

For a short description of this competition, please refer to the call
for participation at

As soon as you decide to take part in this competition, please send an
email to [log in to unmask] to declare your intention to
compete. Early registration is strongly encouraged, so that the
organizing committee is aware of all entrant teams, and can then keep
them informed of any update regarding the organization of the
competition (e.g. deadline extension). It is also worth noting that
taking part in this competition does not require a GECCO registration,
unless a submitted GECCO Companion abstract is accepted for

Please feel free to forward this announcement to any colleagues who
may be interested in this optimization competition.

We are looking forward to your registrations and welcome questions and comments.

Sender:  Manuel López-Ibáñez <[log in to unmask]>
Subject: Call for Papers: Workshop on Understanding Reproducibility in
Evolutionary Computation (Benchmarking@GECCO2021)

1st Workshop on Understanding Reproducibility in Evolutionary Computation

co-organised with

Good Benchmarking Practices for Evolutionary Computation

to be held online as part of

The Genetic and Evolutionary Computation Conference (GECCO 2021),
July 10-14 2021.

Experimental studies are prevalent in Evolutionary Computation (EC), and
concerns about the reproducibility and replicability of such studies has
increased in recent years, following similar discussions in other scientific
fields. In this workshop, we want to raise awareness of the reproducibility
issue, shed light on the obstacles when trying to reproduce results, and
discuss best practices in making results reproducible as well as reporting
reproducibility results.

We invite submissions of papers repeating an empirical study published in a
journal or conference proceedings, either by re-using, updating or
reimplementing the necessary codes and datasets, irrespectively of whether this
code was published in some form at the time or not.

  * The original study being reproduced should not be so recent as to make the
    reproduction attempt trivial. Ideally, we suggest looking at studies that
    are at least 10 years old. However, one of the criteria for acceptance is
    what can the GECCO community learn from the reproducibility study.

  * At least one of the co-authors of the submitted paper should be one of the
    co-authors of the original study. This condition makes sure that the
    reproducibility attempt is a fair attempt at reproducing the original work.

We expect in the submitted paper:

  * Documentation of the process necessary to re-run the experiments. For
    example, you may have to retrieve the benchmark problems from the web,
    downgrade your system or some libraries, modify your original code because
    some library is nowhere to be found, reinstall a specific compiler or
    interpreter, etc.

  * A discussion on whether you consider your paper reproducible, and why you
    think this is the case. If you ran your code with fixed random seeds and you
    have recorded them, you may be able to reproduce identical results. If you
    haven’t recorded the random seeds, you may need to use statistical tests to
    check whether the conclusions still hold. You may even want to try some
    different problem instances or parameter settings to check whether results
    still hold for slightly different experimental settings.

  * Sufficient details to allow an independent reproduction of your experiment
    by a third party, including all necessary artifacts used in the attempt to
    reproduce results (code, benchmark problems, scripts to generate plots or do
    statistical analysis). Artifacts should be made publicly and permanently
    available via Zenodo ( or other data repository or
    submitted together with the paper to be archived in the ACM Digital Library.

In the end, there may be various possible outcomes, and all are acceptable for
a paper: you are unable to run or compile the code, you are able to run the
code but it does not give you the expected results (or no result at all), the
program crashed regularly (before getting results), you do not remember the
parameter settings used, etc. All these are valid conclusions. We care more
about the description of the process, challenges to reproduce results, and the
lessons to be learned, than about whether you have actually been able to
reproduce the study.

Important Dates

   - 11 February 2021: Submissions open
   - 12 April 2021: Submissions deadline
   - 26 April 2021: Acceptance decisions
   - 3 May 2021: Camera-ready papers due and author registration deadline
   - July 10th-14th 2021: Online GECCO conference

Sender: El-ghazali Talbi <[log in to unmask]>
Subject: CFP META'2021@Marrakech

International Conference on Metaheuristics and Nature Inspired Computing
                                                    27-30 Oct 2021
                                            Marrakech, Morocco

META is one of the main event focusing on the progress of the area of
metaheuristics and their applications. As in previous editions,
META’2021 will provide an opportunity to the international research
community in metaheuristics to discuss recent research results, to
develop new ideas and collaborations, and to meet old friends and make
new ones in a friendly and relaxed atmosphere.

All selected long papers will be published in the Springer (SCOPUS, ISI,
DBLP) conference proceedings.
At least, 2 special special issues in ISI and SCOPUS journals are also

META'2021 welcomes presentations that cover any aspects of metaheuristic
research such as new algorithmic developments, high-impact applications,
new research challenges, theoretical developments, implementation
issues, and in-depth experimental studies. META'2021 strives for a
high-quality program that will be completed by a number of invited
talks, tutorials, workshops and special sessions.

The scope of the META conference includes, but is not limited to:
     * Local search, tabu search, simulated annealing, VNS, ILS, …
     * Evolutionary algorithms, swarm intelligence, bio-inspired
algorithms, …
     * Emergent nature inspired algorithms: quantum computing,
artificial immune systems, bee colony, DNA computing, …
     * Quantum computing
     * Parallel and distributed algorithms
     * Decomposition methods
     * Hybrid methods with machine learning, game theory, mathematical
programming, constraint programming, co-evolutionary, …
     * Application to: logistics and transportation, networks,
scheduling, data mining, engineering design, energy, cloud, bio-medical, …
     * Theory of metaheuristics, landscape analysis, convergence,
problem difficulty, very large neighbourhoods, …
     * Multi-objective optimization, bi-level optimization
     * Dynamic optimization, problems with uncertainty, stochastic
optimization, …
     * Parameter tuning (static, dynamic, adaptive, self-adaptive)
     * Hyper-heuristics, cross-domain metaheuristics
     * Software frameworks for metaheuristics and nature inspired computing

Important dates:

- Submission deadline: April 30, 2021
- Notification of acceptance: June 11, 2021


 - Send submissions (articles) to [log in to unmask]
 DO NOT send submissions to the [log in to unmask] address.

 - To subscribe send email to [log in to unmask] containing the
 following text in the body of the message:

   subscribe ec-digest-l <Your Name (up to 4 words)>

 - To unsubscribe send email to [log in to unmask] containing the
 following text in the body of the message:

   unsubscribe ec-digest-l

 - To change your email address, simply unsubscribe the old address
 and subscribe the new one.

 - Send other administrative requests to [log in to unmask]

 End of Evolutionary Computation Digest