Welcome to MESS 2024




MESS 2024 flyer


The 3rd international Metaheuristics Summer School – MESS 2024 – is aimed at qualified and strongly motivated MSc and PhD students; post-docs; young researchers, and both academic and industrial professionals to provide them with an overview on the metaheuristics techniques, and an in-depth analysis of the state-of-the-art. It is a full-immersion four-days course in the beautiful city of Catania whose goal is to offer all participants with a stimulating environment to study and learn advanced concepts and recent research in the fields of Metaheuristics, Optimization, Machine Learning and Artificial Intelligence, in general.

The main theme of this third edition is focused on “Automated Deep Learning Meets High-Performance Computing”, that is how these research areas may interact and affect each other in order to develop reliable and robust solving methodologies for Big Data analysis, and data-driven problems.

The courses will be taught by world-renowned experts in the field and will involve a total of 36-40 hours of lectures, therefore according to the academic system, all PhD and master students attending the summer school will can get 8 ECTS points. To this end, during the school the students will tackle homework, and/or project development for final evaluation. Furthermore, during the four days, the participants will have plenty of opportunities to debate and work with leaders in the field, benefiting from direct interaction and discussions in a friendly environment.

MESS 2024 also offers the possibility to all attendants to present their recent results and/or their works in progress through interaction with their scientific peers, in a professional and constructive environment.

As carried out in the past editions, also in this edition all participants will be involved in the “Metaheuristics Competition Race”, where each of them, individually or divided in working groups (no more than 3 people), will develop a metaheuristic solution for a given problem presented during the school. The top three of the competition ranking will receive the MESS 2024 award and will be involved in the writing of a manuscript dedicated to the competition that, afterwards, will be submitted to an international journal for possible publication. In addition, the remaining best ten in the ranking (excluding the top three, of course) will be invited to report their work in a manuscript which will be published in the special MESS 2024 Volume of the AIRO Springer Series.

See the publications of the previous Metaheuristics Competitions:

  • MESS 2020+1 Metaheuristics Competition [link];
  • MESS 2018 Metaheuristics Competition [link].

 
 

 

 


Lecturers

   

   

Thomas Bartz-Beielstein

TH Köln, Germany

Marco Gori

University of Siene, Italy

Aaron Klein

AWS Research Berlin, Germany

 

 

Eric Medvet

University of Trieste, Italy

David Salinas

University of Freiburg, Germany

El-Ghazali Talbi

University of Lille 1, France

 



 



 



 

 

Courses

 

Thomas Bartz-Beielstein

 

Lecture 1

Introduction to Surrogate-Based Hyperparameter Tuning in Classical Machine Learning, Online Machine Learning, and Deep Learning

   

Lecture 2

Exploring Hyperparameter-Tuning Applications: Optimizing scikit-learn, River, and PyTorch with SPOT and Other Tuning Techniques

Marco Gori

 

Lecture 1

HAMILTONIAN LEARNING: Foundations of Lifelong Learning

   

   

Lecture 2

HAMILTONIAN LEARNING: Case studies
   
   
   

Aaron Klein

 

Lecture 1

Introduction into AutoML:
Hyperparameter Optimization

   
   

   

Lecture 2

Introduction into AutoML:
Neural Architecture Search


   

Eric Medvet

 

Lecture 1

Metaheuristics for policy search in continuous control

   

Lecture 2

Beyond policy search: optimizing body and brain of simulated robotic agents
   
   

David Salinas

 

Lecture 1

Multi-objective optimization
   

   

Lecture 2

Speeding-up Hyperparameter Optimization with transfer and meta learning
   
   

El-Ghazali Talbi

 

Lecture 1

Metaheurististics for neural architecture search and hyperparameter optimization

   

Lecture 2

Parallel and multi-objective metaheuristics for neural architecture search and hyperparameter optimization
   







 

 

Scientific Program

Download the Program



   

 

 

Oral Presentations

The Metaheuristics Summer School aims to provide a stimulating environment for all participants, where they may interact and exchange ideas with their scientific peers in a friendly and constructive way. To this end to all participants are offered the opportunity to present their research activities on the Metaheuristics by an oral presentation.

During the application, all participants, by filling the relative section, may submit an abstract of their recent results, or works in progress, to be presented in a dedicated session as an oral presentation. The abstract must be submitted before March 16, and must be written in English. Each abstract will be reviewed by the Workshop Chairs and/or Scientific Committee. All abstracts, in any format, will be included in an electronic book of MESS 2024, which will be both put online on the website and provided to all participants during the school.

 

For more information send an email to: mess.school @ ANTs-lab.it

 

 

Abstract Booklet

Download the Abstract Booklet



   

   

 

 

Metaheuristics Competition

Trigger Arc TSP

The Trigger Arc Travelling Salesman Problem (TATSP) is defined as the problem of identifying in a graph G a Hamiltonian cycle of minimum cost considering relations defined between arcs. For each arc a in the graph, relations with other arcs can be defined, which act as triggers. If the trigger arc is traversed, the cost of the arc a is reset. Only the last trigger arc encountered before reaching the arc a is considered. The key concept in this problem is the definition of a trigger arc. If the relationship (t,a) is defined for arcs t and a, then arc t will act as a trigger for arc a, meaning that traversing arc t will result in the assignment of a new cost to arc a if a has not yet been traversed.

 

All information needed, included its formulation, can be found at the following link:
https://fourclicks.eu/fck/mess2024/frontend/#

 

 

Competition Description

 

  • Participants will need to form groups consisting of up to 3 people.
  • Participants are expected to develop the most suitable metaheuristic algorithms to solve the given problem.
  • The created solutions must be uploaded to the portal.
  • The score for each instance will be calculated based on the objective function value and the submission date of the solution.
  • The leaderboard will be created based on the sum of the scores of individual instances.
  • During the three months of the competition, new instances will be released, usually created with different generators. The use of heterogeneous instances is intended to ensure that the developed techniques are adaptable and flexible.
  • Participants can update their submitted solutions at any time until the end of the competition.

 

 

Presentation



   

   

 

 

Important Dates

Application Deadline: April 16, 2024

Notification of acceptance: May 5, 2024

Early Registration: by May 20, 2024

Late Registration: from May 21, 2024




 

 

 

Application

The applications are closed

   

MESS 2024 is addressed to motivated, qualified, and pre-selected candidates interested in the topics proposed by the school. Thus PhD students; Post-Docs; and young researchers (both academic and industrial) as well as advanced graduate and undergraduate students are welcomed and encouraged to apply.

Applications to attend MESS 2024 must be received before 16th April 2024 filling the module below. All accepted applicants will be notified by 25th April 2024 in accordance with the application time.

  

SCHOOL FEE:


   Early registration (by May 20, 2024): € 550;
   Late registration (from May 21, 2024): € 700.

  

The registration fee covers, respectively:

  • all course sessions,
  • lesson materials,
  • oral presentations,
  • all scientific activities,
  • coffee breaks,
  • welcome party,
  • guided tour,
  • social dinner,
  • workshop,
  • VAT.

Note that extra costs for bank wire transfers or credit card service are not covered by the school.

Those who would like to give an oral presentation on their recent results, or works in progress may include Title and Abstract in the relative application section.

 

 

Application Form



-

Oral Presentation:

Consent: please review the italicized text below and then click the checkbox to indicate your consent:

“As the submitting author, I attest the integrity of the data and experimental methodologies used. I have given all abstract authors the appropriate credit and secured the necessary permission from the institution(s) and authors listed.

I grant MESS 2024 the right to publish my abstract on their website and secure mobile app, which are provided to registered attendees, and after the conference, make my abstract available by other interested persons.”



 

 

 

Directors

Salvatore Greco

University of Catania, Italy

Mario Pavone

University of Catania, Italy

Andrea Schaerf

University of Udine, Italy

 

 

El-Ghazali Talbi

University of Lille 1, France

Daniele Vigo

University of Bologna, Italy

 

 

 

 

 

 

Metaheuristics Competition Chairs

 

Raffaele Cerulli

University of Salerno, Italy

Sara Ceschia

University of Udine, Italy

 

 

Carmine Cerrone

University of Genoa, Italy

Andrea Schaerf

University of Udine, Italy

 

 

 

Local Organizing Committee

Vincenzo Cutello (Chair), University of Catania, Italy
Claudia Cavallaro, University of Catania, Italy
Carolina Crespi, University of Catania, Italy
Angelo Cavallaro, University of Catania, Italy
Angela Pappalardo, ANTs Research Laboratory, Italy
Antonino Di Stefano, ANTs Research Laboratory, Italy
Alessandro Vitale, ANTs Research Laboratory, Italy
Alessio Mezzina, University of Catania, Italy
Attilio Di Natale, University of Catania, Italy

Publicity Chair

Francesco Zito, University of Catania, Italy
Stephan Petrone, ANTs research laboratory, Italy

 

Sponsorship & Patronage

 

Sponsorship




   

   

                   


   

   

                   


               

   

   

Patronage

   

 

 

 

 

 

 

Venue

MESS 2024 will take place at the Department of Economics and Business of the University of Catania. The Department is located in the Palazzo delle Scienze, a beautiful and historical building situated in the city center, in Corso Italia 55.

 

Department of Economics and Business

Palazzo delle Scienze

Corso Italia 55, 95129 Catania, Italy

 

DEI    DEI    DEI

 

 

 

Accommodation

Catania is an ancient port city on Sicily’s east coast. It sits at the foot of Mt. Etna, an active volcano with trails leading up to the summit. The city’s wide central square, Piazza del Duomo, features the whimsical Fontana dell’Elefante statue and the richly decorated Catania Cathedral. In the southwest corner of the square, La Pescheria (a weekday fish market) is a rowdy spectacle surrounded by seafood restaurants.

It is a modern, sparkling city, always full of vitality; indeed it is recognized as the liveliest city in Sicily. The black and white of its buildings contrasts with the blue of the sky and sea; and at night, its squares and avenues are populated with young people who love to have fun in the hundreds of clubs that fill the historic center.

Catania offers a wide choice of accommodation from small family-run guest houses to traditional B&Bs; from Vacation Rentals to modern Hotels. Here you can also find a broad selection of excellent restaurants and cosy trattorias where you can taste the genuine, true flavours of the authentic sicilian cuisine.

 

 

 

 

Contact US

ANTs Research Laboratory

+39 338 4342147
mess.school@ANTs-lab.it
Secretary:
Eng. Angela Pappalardo
via SS. Crocifisso 66
I-95024, Acireale (CT), Italy



Follow us on Facebook