Previous Next
Scroll down to
explore more

LION17 Scope

The 17th Learning and Intelligent OptimizatioN Conference

The central theme of this conference is ML to OR pipelines. When integrating multiple data science pipelines, significant challenges arise, especially when considering predictive and prescriptive analytics. Central challenges here include loss functions for training ML models that will be used downstream by optimization approaches and actively taking uncertainty into account in optimization models. This joins with classic LION themes, such as determining appropriate optimization methods through expensive algorithm configuration and parameter tuning , implementing intelligent learning schemes for learning from past algorithm behavior to improve performance in the future, hybridizing different algorithms (evolutionary, etc.) to achieve robust and effective performance, and so on.

This meeting continues the successful LION series (8.Gainesville, 9.Lille, 10.Ischia, 11.Nizhny, 12.Kalamata, 13.Chania, 14.Athens, 15.Athens, 16.Milos), explores the intersections and uncharted territories between machine learning, artificial intelligence, mathematical programming and algorithms for hard optimization problems.

*IBM is a co-organizer

Important dates

All deadlines are Anywhere on Earth (AoE = UTC-12h).

  • September 15, 2022, Special Sessions proposals submission deadline
  • September 21, 2022, Special Sessions notification of acceptance
  • January 23, 2023February 1, 2023, abstract submission (for full paper submission)
  • February 1, 2023February 8, 2023, full paper submission
  • March 8, 2023March 15, 2023, author notification
  • March 1, 2023March 17, 2023, presentation only abstracts
  • April 4, 2023, registration opens
  • April 21, 2023, early registration closes
  • May 1, 2023, conference pre-proceedings
  • May 13, 2023, late registration closes
  • May 31, 2023 proposals to organize the future editions of LION
  • June 4-8, 2023, conference at Nice, France



Registration is closed

Call for Papers

Topics of Interest

The 17th Learning and Intelligent Optimization conference is an interdisciplinary conference that brings together researchers in operations research, machine learning and related fields. We welcome contributions on all LION themes, i.e., topics containing exact and heuristic optimization involving learning or intelligent techniques, such as determining and applying appropriate optimization methods, expensive algorithm configuration, intelligent learning schemes for learning from past algorithm behavior, hybrid algorithms (evolutionary, etc.) to achieve robust and effective performance. Also, we solicit contributions on this year's special theme "ML and OR pipelines".

We invite submissions presenting new and original research on all topics related and relevant to the conference. All papers will be reviewed in a double-blind process and accepted papers will be presented at the conference.


Papers accepted into the LION17 proceedings will be published in Lecture Notes in Computer Science (LNCS). Submitted papers can have a maximum length of 15 pages (including references) and should use the LNCS template, which can be found here: Papers must be submitted as a PDF in English.

Please note that concurrent submissions are not allowed, and that a unique author of each accepted paper must register for conference for the paper to be included in the proceedings.


Submit at:

The site will start accepting submissions on November 1, 2022.

The submission of presentation only abstracts through the link above will be possible from 22 January 2023

Presentation only abstracts: We also offer the opportunity to submit presentation-only abstracts of up to 2000 characters regarding ongoing or recently published work. These can be submitted as PDF but do not need to adhere to the LNCS format.



17:30 - 20:00Registration
18:00 - 20:00Welcome Social Event
(Extra registration required)
8:00 - 17:00Registration
8:30 - 8:45Opening
8:45 - 9:50(5088) Learning to Prune Electric Vehicle Routing Problems
James Fitzpatrick, Deepak Ajwani and Paula Carroll
(4002) Repositioning Fleet Vehicles: a Learning Pipeline
Augustin Parjadis, Quentin Cappart, Quentin Massoteau and Louis-Martin Rousseau
(9167) Dynamic Police Patrol Scheduling with Multi-Agent Reinforcement Learning
Songhan Wong, Waldy Joe and Hoong Chuin Lau
Session chair: Hoong Chuin Lau
9:50 - 10:20Break
10:20 - 11:15(7001) Job Shop Scheduling via Deep Reinforcement Learning: a Sequence to Sequence approach
Giovanni Bonetta, Davide Zago, Rossella Cancelliere and Andrea Grosso
(3715) Predict, Tune and Optimize for Data-Driven Shift Scheduling with Uncertain Demands
Michael Römer, Felix Hagemann and Till Porrmann
(9999) Learning Heuristics for Combinatorial Optimization Problems with Deep Neural Networks
André Hottung
Session chair: Michael Römer
11:15 - 11:30Break
11:30 - 12:30Invited Keynote: Process Mining for Optimization and Optimization for Process Mining
Wil van der Aalst (RWTH Aachen, Celonis)
Chair: Kevin Tierney
12:30 - 13:50Lunch sponsored by Gurobi
13:50 - 14:55(2111) Real-world process discovery from low-level event data
Stephane Cholet, Franck Lefebure, Cecile Thuault and Gerald Christophe
(568) Anomaly Classification to Enable Self-Healing in Cyber Physical Systems using Process Mining
Uphar Singh, Deepak Gajjala, Rahamatullah Khondoker, Harshit Gupta, Ayush Sinha and O.P. Vyas
(1669) Explaining the Behavior of Reinforcement Learning Agents using Association Rules
Zahra Parham, Vi Tching de Lille and Quentin Cappart
Session chair: O.P. Vyas
14:55 - 15:30Break
15:30 - 17:05(4555) Relational Graph Attention-based Deep Reinforcement Learning:
An Application to Flexible Job Shop Scheduling with Sequence-dependent Setup Times

Amirreza Farahani, Martijn Van Elzakker, Laura Genga, Pavel Troubil and Remco Dijkman
(4836) Experimental Digital Twin for Job Shops with Transportation Agents
Aymen Gannouni, Luis Felipe Casas Murillo, Marco Kemmerling, Anas Abdelrazeq and Robert H. Schmitt
(5339) A matheuristic approach for electric bus fleet scheduling
Raka Jovanovic, Sertac Bayhan and Stefan Voss
(4466) Expert-Iteration for Combinatorial Optimization
Arnaud Sors, Darko Drakulic, Florian May, Sofia Michel and Jean-Marc Andreoli
(333) Hyperparameter Optimization and Weight Initializations: A Performance Analysis for Deep Learning Models
Nicki Lena Kämpf
(8862) Reinforcement Learning for routing in mixed-shelves warehouses
Laurin Luttmann and Lin Xie
Session chair: Raka Javanovic
8:45 - 9:50(9680) Unleashing the potential of restart by detecting the search stagnation
Yoichiro Iida, Tomohiro Sonobe and Mary Inaba
(3301) Heuristics selection with ML in CP Optimizer
Hugues Juille, Renaud Dumeur and Paul Shaw
(4238) Towards Tackling MaxSAT by Combining Nested Monte Carlo with Local Search
Hui Wang, Abdallah Saffidine and Tristan Cazenave
Session chair: Tristan Cazenave
9:50 - 10:20 Break
10:20 - 11:15(8347) Learn, Compare, Search: One Sawmill’s Search for the Best Cutting Patterns Across And/or Trees
Marc-André Ménard, Michael Morin, Mohammed Khachan, Jonathan Gaudreault and Claude-Guy Quimper
(2957) GPU for Monte Carlo Search
Lilian Buzer and Tristan Cazenave
(5711) Enabling Flexible Grid Management through Intelligent Demand Response
Scheduling and Neural Network-based NILM in Residential Microgrids

Mohamed Saâd El Harrab and Michel Nakhla
Session chair: Roberto Battiti
11:15 - 11:30Break
11:30 - 12:30Invited Tutorial: Predict-then-Optimize: a Tour of the State-of-the-art using PyEPO
Elias Khalil (University of Toronto)
Chair: Michael Römer
12:30 -13:50Lunch
13:50 - 15:20(613) Hyper-box classification model using mathematical programming
Georgios I. Liapis and Lazaros G. Papageorgiou
(1175) Hierarchical Machine Unlearning
Zhu Hongbin, Xia Yuxiao, Li Yunzhao, Li Wei, Liu Kang and Gao Xianzhou
(5489) Surrogate Membership for Inferred Metrics in Fairness Evaluation
Melinda Thielbar, Serdar Kadioglu, Chenhui Zhang, Rick Pack and Lukas Dannull
(5963) Discovering Explicit Scale-Up Criteria in Crisis Response with Decision Mining
Britt Lukassen, Yingqian Zhang, Laura Genga and Michiel Rhoen
Session chair: Laurens Bliek
15:20 - 15:45Break
15:45 - 17:00(7213) Generating a Graph Colouring Heuristic with Deep Q-Learning and Graph Neural Networks
George Watkins, Giovanni Montana and Juergen Branke
(1766) Deep Randomized Networks for Fast Learning
Richárd Rádli and László Czúni
(2963) Learning the Bias Weights for Generalized Nested Rollout Policy Adaptation
Julien Sentuc, Farah Ellouze, Jean-Yves Lucas and Tristan Cazenave
(4488) Optimizing Product Offerings with Deep Reinforcement Learning
Chaher Alzaman
Session chair: Yingqian Zhang
19:00- 21:00Conference Banquet
(Extra registration required)
8:45 - 9:50(2087) Generative models via Optimal Transport and Gaussian Processes
Antonio Candelieri, Andrea Ponti and Francesco Archetti
(3921) On Learning When to Decompose Graphical Models
Aleksandra Petrova and Javier Larrosa
(1131) Sensorimotor Learning with Stability Guarantees via Autonomous Neural Dynamic Policies
Dionis Totsila, Konstantinos Chatzilygeroudis, Dimitrios Kanoulas and Ioannis Hatzilygeroudis
(3745) Automated packing systems
David Alvarez Martinez, Daniel Giraldo, Ana Maria Montes Franco,Daniel Cuellar-Usaquen, German Fernando Pantoja Benavides,
Ruben Iván Bolaños, Carlos Rodriguez, Valentina Bedoya and Juan Martínez
Session chair: Konstantinos Chatzilygeroudis
9:50 - 10:20Break
10:20 - 11:15(728) A leak localization algorithm in water distribution networks using
probabilistic leak representation and optimal transport distance

Andrea Ponti, Ilaria Giordani, Antonio Candelieri and Francesco Archetti
(3602) An Error-Based Measure for Concept Drift Detection and Characterization
Antoine Bugnicourt, Riad Mokadem, Franck Morvan and Nadia Bebeshina
(3727) Machine Learning for Combinatorial Optimisation of Partially-Specified Problems: Regret Minimisation as a Unifying Lens
Stefano Teso, Laurens Bliek, Andrea Borghesi, Michele Lombardi, Neil Yorke-Smith, Tias Guns and Andrea Passerini
Session chair: Francesco Archetti
11:15 - 11:30Break
11:30 - 12:30Invited Keynote: Operations Research + Machine Learning for the design of future offshore wind farms
Martina Fischetti (European Commission Research Center)
Chair: Dario Pacino
12:30 - 13:50 Lunch
13:50 - 15:20(3464) Model-based feature selection for neural networks: A mixed-integer programming approach
Shudian Zhao, Calvin Tsay and Jan Kronqvist
(5732) The BeMi Stardust: a Structured Ensemble of Binarized Neural Networks
Ambrogio Maria Bernardelli, Simone Milanesi, Stefano Gualandi and Hoong Chuin Lau
(7633) Multi-Task Predict-then-Optimize
Bo Tang and Elias Khalil
(8137) Improving Subtour Elimination Constraint Generation in Branch-and-Cut Algorithms for the TSP with Machine Learning
Thi Quynh Trang Vo, Viet Hung Nguyen, Paul Weng and Mourad Baiou
Session chair: Elias Khalil
15:20 - 15:45Break
15:45 - 16:45(8888) Algorithm Configuration in the UPF: Exploiting capabilities of Selector
Dimitri Weiß, Elias Schede, Kevin Tierney
(2088) Multi-Agent Reinforcement Learning for Strategic Bidding in Two Stage Electricity Markets
Francesco Morri, Hélène Le Cadre, Luce Brotcorne and Pierre Gruet
(4968) Learning with Neural Networks to solve Resource Leveling Problems
Lena Wohlert and Jürgen Zimmermann
(7091) CLS-Luigi: A Framework for Automated Synthesis and Execution of Decision Pipelines
Anne Meyer, Jan Bessai, Hadi Kutabi and Daniel Scholtyssek
(1371) The Dynamic RORO Stowage Planning Problem
Alastair Main, Filipe Rodrigues and Dario Pacino
(9237) Metaheuristics for optimal localisation of weather radars network over an airspace
Abdessamed Mogtit, Redouane Boudjemaa and Mohand Lagha
Session chair: André Hottung
8:45 - 10:00(1124) Bayesian Optimization for Function Compositions with Applications to Dynamic Pricing
Kunal Jain, Prabuchandran K.J. and Tejas Bodas
(1173) A Bayesian optimization algorithm for constrained simulation optimization problems with heteroscedastic noise
Sasan Amini and Inneke Van Nieuwenhuyse
(4133) Bayesian Decision Trees Inspired from Evolutionary Algorithms
Efthyvoulos Drousiotis, Alexander Phillips, Paul Spirakis and Simon Maskell
(4570) Managing Perishable Inventory with Distributional Forecasts
Mark Velednitsky and Philip Cerles
Session chair: Antonio Candelieri
10:00 - 10:30Break
10:30 - 11:30Invited Tutorial: SAT-based Applications with OptilLog
Carlos Ansótegui (University of Lleida)
Chair: Kevin Tierney
11:30 - 12:15(864) Fast and Robust Constrained Optimization via Evolutionary and Quadratic Programming
Konstantinos Chatzilygeroudis and Michael Vrahatis
(9199) Analysis of Heuristics for Vector Scheduling and Vector Bin Packing
Lars Nagel, Nikolay Popov, Tim Suess and Ze Wang
Session chair: Ioannis Chatzilygeroudis
12:15 - 12:45Closing
12:45 - 14:00Lunch

Book of abstracts

A book of abstracts can be downloaded.

Keynotes and Tutorials

Keynote 1: Wil van der Aalst (RWTH Aachen University, Celonis)

Process Mining for Optimization and Optimization for Process Mining

Abstract: Introduced over 20 years ago, process mining is rapidly becoming the standard way to diagnose and improve processes based on event data hidden in any information system. The process mining discipline is moving from simple 2D processes using a single case notion to 3D processes involving multiple object types. It is possible to discover object-centric process models from event data and check the conformance of such models to identify compliance and performance problems. Many of the larger organizations are already using process-mining tools (e.g., Celonis) to improve processes. Moreover, process mining enables the application of machine learning and optimization in a wide range of settings (production, healthcare, logistics, administration, finance, energy, etc.). The keynote will focus on the interconnection between process mining and optimization. Because process mining is data-driven, the focus is on finding, diagnosing, and predicting problems in operational processes. However, in settings with limited resources and time constraints, one needs to make decisions to improve processes. Also, next to events that have happened, there are events that are planned or scheduled. Process mining helps to automatically generate optimization problems in ill-defined settings. Moreover, optimization techniques ranging from AI planning to mixed integer linear programming can be used to perform process mining tasks. A well-known example is the computation of alignments for conformance checking using the marking equation and the discovery of control-flow constructs in process models. Here, advances in mathematical programming are used to speed-up process mining algorithms. Given the interesting links and huge application potential, the keynote will encourage participants of the 17th Learning and Intelligent Optimization conference to work on the intersection of both fields.

Bio: Wil van der Aalst is a full professor at RWTH Aachen University, leading the Process and Data Science (PADS) group. He is also the Chief Scientist at Celonis, part-time affiliated with the Fraunhofer FIT, and a member of the Board of Governors of Tilburg University. He also has unpaid professorship positions at Queensland University of Technology (since 2003) and the Technische Universiteit Eindhoven (TU/e). Currently, he is also a distinguished fellow of Fondazione Bruno Kessler (FBK) in Trento, deputy CEO of the Internet of Production (IoP) Cluster of Excellence, and co-director of the RWTH Center for Artificial Intelligence. His research interests include process mining, Petri nets, business process management, workflow automation, simulation, process modeling, and model-based analysis. Many of his papers are highly cited (he is one of the most-cited computer scientists in the world and has an H-index of 170 according to Google Scholar with over 130,000 citations), and his ideas have influenced researchers, software developers, and standardization committees working on process support. He previously served on the advisory boards of several organizations, including Fluxicon, Celonis, ProcessGold/UiPath, and aiConomix. Van der Aalst received honorary degrees from the Moscow Higher School of Economics (Prof. h.c.), Tsinghua University, and Hasselt University (Dr. h.c.). He is also an IFIP Fellow, IEEE Fellow, ACM Fellow, and an elected member of the Royal Netherlands Academy of Arts and Sciences, the Royal Holland Society of Sciences and Humanities, the Academy of Europe, the North Rhine-Westphalian Academy of Sciences, Humanities and the Arts, and the German Academy of Science and Engineering. In 2018, he was awarded an Alexander-von-Humboldt Professorship.

Keynote 2: Martina Fischetti (European Commission Research Center)

Operations Research + Machine Learning for the design of future offshore wind farms

Abstract: Sustainability is a key focus in our society that is today working to change towards a greener future. Wind energy, in particular, is attracting always more attention as source of renewable energy. In this picture, Vattenfall is working towards the ambitious goal of becoming fossil free within one generation. To achieve this goal, innovation (and optimization!) is of key importance.
This talk presents how Vattenfall is using advanced operations research and analytics for designing cheaper and more profitable offshore wind farms. The talk will focus on the design phase of offshore wind farms, explain in details the optimization challenges faced by companies as Vattenfall. In particular, we will focus on the Offshore Wind Farm Design problem, that is the task of deciding how to position turbines offshore in order to increase the overall farm production and reduce costs. This task is particularly challenging due to the interference effects among turbines, due to the stochasticity of wind and due to the high dimensionality of the problem in real applications. Mixed Integer Programming models and other state-of-the-art optimization techniques have been developed to solve this problem. These tools are nowadays fully deployed in Vattenfall and used for the design of all offshore wind farms. They have been used, for example, for the design of Hollandse Kust Zuid in the Netherlands, which is the first offshore wind farm ever constructed without any subsidies. This is a huge milestone for the whole wind energy business.
These advanced optimization tools allowed Vattenfall to think out of the box, take more informed decision and perform different what-if-analyses. In particular, we can foresee the number of what-if analyses to quickly grow in the future. Therefore we have looked into Machine Learning techniques.
In the specific, we propose a combination of Mathematical Optimization and Machine Learning to estimate the value of optimized solutions. We investigate if a machine, trained on a large number of optimized solutions, could accurately estimate the value of the optimized solution for new (unseen) instances. This research question could be of general interest for the OR community, but we focus on the wind farm layout application in our research. Given the complexity of the wind farm layout problem and the big difference in production between optimized/non optimized solutions, it is not trivial to understand the potential value of a new site without running a complete optimization. This could be too time consuming if a lot of sites need to be evaluated, therefore we propose to use Machine Learning to quickly estimate the potential of new sites (i.e., to estimate the optimized production of a site without explicitly running the optimization). To do so, we trained and tested different Machine Learning models on a dataset of 3000+ optimized layouts found by the optimizer. Our results show that Machine Learning is able to efficiently estimate the value of optimized instances for the offshore wind farm layout problem.

Bio: Martina Fischetti worked as lead engineer in Vattenfall BA Wind until November 2021, specialized in operational research (OR). She holds M.Sc. degrees from the University of Padova (March 2014) and the University of Aalborg (June 2014) in Automation Engineering. In March 2018, she finished her Industrial PhD in OR at Technical University of Denmark in collaboration with Vattenfall, entitled Mathematical Programming Models and Algorithms for Offshore Wind Park Design. Her PhD work on the optimization of wind farm design and cable routing has been awarded various international prizes, such as the Best Industrial PhD from Innovation Fund Denmark (2019), EURO Doctoral Dissertation Award (2019), Glover-Klingman Prize (2018), AIRO Best Application Paper award (2018), the Best Student Paper Award ICORES (2017), and finalist positions at the EURO Excellence in Practice award (2018) and the prestigious INFORMS Franz Edelman award (2019). She was also selected as role model for young women in OR by the EURO WISDOM forum in 2021. She currently works as senior researcher – policy analyst at the Joint Research Center of the European Commission in Seville, Spain, where she applies Operations Research to European transport challenges.

Tutorial 1: Carlos Ansótegui (University of Lleida)

SAT-based Applications with OptilLog

Bio: Carlos Ansótegui is a professor of the computer science department at the University of Lleida (UdL). He leads the Logic & Optimization Group (LOG), which brings together researchers from UdL and Universitat Politècnica de Catalunya (UPC). LOG is focused on the design of efficient solving techniques for combinatorial optimization problems through Satisfiability-based solving approaches, efficient modeling, and automatic configuration.

Abstract: OptiLog is a Python framework for the rapid prototyping of SAT-based systems. OptiLog includes functionality for loading and creating formulas, dynamic loading of incremental SAT solvers with support for external libraries, modules for modelling problems into Non-CNF format with support for Pseudo Boolean constraints, for tuning, evaluating and parsing the results of applications. All these enhancements allow OptiLog to become a swiss knife for SAT-based applications in academic and industrial environments. During this talk, we will glimpse what we can do with OptiLog.

Tutorial 2: Elias Khalil (University of Toronto)

Predict-then-Optimize: a Tour of the State-of-the-art using PyEPO

Bio: Elias B. Khalil is the Scale AI Research Chair in Data-Driven Algorithms for Modern Supply Chains, an Assistant Professor of Industrial Engineering at the University of Toronto, and a Faculty Affiliate of the Vector Institute. Prior to that, he was the IVADO Postdoctoral Scholar at Polytechnique Montréal. Elias obtained his Ph.D. from the College of Computing at Georgia Tech where he was an IBM Ph.D. Fellow in 2016. His research interests are in the integration of machine learning and discrete optimization.

Abstract: In deterministic optimization, it is typically assumed that all parameters of the problem are fixed and known. In practice, however, some parameters may be a priori unknown but can be estimated from historical data. A typical predict-then-optimize approach separates predictions and optimization into two stages. Recently, end-to-end predict-then-optimize has become an attractive alternative. In this talk, I will provide an overview of the predict-then-optimize problem as well as a suite of end-to-end learning methods that have been developed independently in the operations research and machine learning communities. To ground the discussion, I will use the PyEPO package (, a PyTorch-based end-to-end predict-then-optimize library in Python. PyEPO (pronounced like "pineapple" with a silent "n") is the first generic tool for linear and integer programming with predicted objective function coefficients. It provides four base algorithms: a convex surrogate loss function from the seminal work of Elmachtoub & Grigas (2021), the differentiable black-box solver approach of Vlastelica et al. (2019), and two differentiable perturbation-based methods from Berthet et al. (2020). PyEPO provides a simple interface for the definition of new optimization problems, the implementation of state-of-the-art predict-then-optimize training algorithms, the use of custom neural network architectures, and the comparison of end-to-end approaches with the two-stage approach. PyEPO enables the first comprehensive comparison of these methods on problems such as Shortest Path, Multiple Knapsack, and the Traveling Salesperson Problem, including on an image dataset. The empirical insights therein could guide future research. Joint work with Ph.D. student Bo Tang at the University of Toronto.

Special Sessions

In addition to submissions about general LION themes, we also welcome submissions related to one of our special sessions. The special sessions will be part of the regular conference and are subject to the same peer-review as all other submissions.

Special Session 1: Responsible AI in practice

Organizer : Hendrik Baier1, Laurens Bliek1, Zaharah Bukhsh1, Isel Grau1, Yingqian Zhang1
1Eindhoven University of Technology, Eindhoven, The Netherlands
Abstract: Artificial Intelligence is increasingly used in real-world applications, offering countless opportunities for businesses, health, and education. Besides opportunities, AI systems can also bring many unexpected problems and unintended consequences in practice, which has recently led to a new focus on trustworthy and responsible AI. It is important to develop and deploy AI algorithms that are not only highly performant, but also, for example, fair, unbiased, transparent, explainable/interpretable, robust, accountable, and sustainable, and that protect the privacy of users. However, each of these possible requirements for responsible AI comes with its own set of challenges in implementation and evaluation. Furthermore, once implemented and applied to different domains, their context and interaction can create new trade-offs that need to be taken into account. This workshop aims at gathering researchers who are interested in responsible AI solutions in practice – with a particular focus on human-focused evaluation, practical implementation, or the consideration of new challenges that arise from transferring responsible AI out of the lab and into the real world.

Special Session 2: Data-Driven Optimization with Business Process Mining

Organizer : Om Prakash Vyas1, Jerome Geyer-Klingeberg 2
1Indian Institute of Information Technology, Allahabad , India
2Celonis, Munich, Germany
Abstract: Process Mining, positioned at the interface between Process Science and Data Science, combines event data with process models and intends to gain insights, identify bottlenecks, predict problems, and optimize organizational processes. Process Mining, already being used for high-volume processes in large organizations, will soon become the ‘new normal’ for smaller organizations and processes with few cases as well.
Despite a huge surge in researching endeavors in Process Discovery, Conformance Checking, and Model Enhancement, positioning them as three verticals of Process Mining, there are a number of research challenges that need to be overcome to realize the vision of data driven optimization of business processes. The optimization paradigm in the process mining context are being explored at following levels:
1. When it comes to creating process models, event logs generated by process-oriented information systems are treated as a critical resource. Conformance checking can be formulated as an optimization problem with the model and log repair. Thus, conformance checking corresponds to solving optimization problems that grow exponentially in the size of the model and the length of traces in the event log
2. Optimization metaheuristics have also been widely applied in the context of automated process discovery, with the goal of gradual discovery and advancement of process models to achieve a tradeoff between accuracy and simplicity. The most notorious of these approaches are those based on evolutionary (genetic) algorithms. However, several other metaheuristics have been researched, such as Imperialist competition algorithms, swarm particle optimization, and simulated glow in this context
3. Data ingestion from diverse source systems is supported by AI, which allows to identify and customize structured and unstructured data from various sources. Thus, various optimization techniques can be used to improve the performance of the data transformation discovery techniques in the context of the synthesis of routine specifications.
With rapidly growing applications in this special session invites original unpublished research contributions that demonstrate current findings in the area of application of data science and optimization techniques for process mining, with special reference to algorithms for process discovery, conformance checking, and process model enhancement.

Special Session 3: Learning and Intelligent Optimization for Physical Systems

Organizer: Michael Vrahatis1, Konstantinos Chatzilygeroudis1
1Department of Mathematics, University of Patras, Greece
Abstract: Several critical challenges arise when operating with physical systems contrary to theoretical models, simulated environments, or static datasets. Firstly, reducing the up-time of experimenting with the systems is essential. Experimenting extensively on a physical system might lead to hardware failures that are expensive to replace. Secondly, the algorithm should never produce behaviors that might harm the humans around it or the system itself (e.g., we do not want to break a robot that costs 2M euros). Therefore, to develop effective Machine Learning or Intelligent Optimization methods on physical systems, one has to consider the above challenges during the process of designing the algorithms. Learning and data-driven methods can learn very complex models/controllers and improve over time which is useful when operating with physical systems. However, such methods require a prohibited amount of samples to work reliably, and providing formal guarantees on the obtained solutions is challenging. On the other hand, traditional mathematical optimization is more often used in physical systems since it can operate with no or little data and provide solid theoretical foundations, but it is not easy to make an algorithm that can improve the performance over time. This special session welcomes submissions on "Learning and Intelligent Optimization for Physical Systems", where the goal is to find novel methods that effectively combine data-driven/ML approaches with mathematical optimization to solve tasks on physical systems. Examples are robot learning for control, sensors, embedded systems/mobile phone algorithms, real-time systems/applications, and human-computer interaction.

Special Session 4: Combinatorial and Integer Optimization Layers in Deep Learning

Organizer: Elias B. Khalil1, Bistra Dilkina2, Axel Parmentier1
1Industrial Engineering, University of Toronto, Canada
2Computer Science, University of Southern California, USA
3Ecole Nationale des Ponts et Chaussées, France
Abstract: A much closer integration of combinatorial and integer optimization with machine learning (ML) has been advocated in the last few years. For one, combinatorial problems such as matching, subset selection, clustering, etc., are useful inductive biases for machine learning tasks in computer vision and natural language processing. On the other hand, data-driven optimization can benefit from machine learning models to construct “accurate” models for optimal decision-making. The key challenge is that combinatorial problems typically resist the differentiability requirement of end-to-end deep learning, and may also require large solution times. This session will focus on methodological advances towards these challenges, efficient and open-source software frameworks, and applications.

Special Session 5: Bayesian Optimization and (Machine) Learning: solutions, applications, challenges, and perspectives

Organizer: Antonio Candelieri1, Francesco Archetti1
1University of Milano-Bicocca, Italy
Abstract: Bayesian Optimization (BO) is considered as a powerful learning-based algorithm to globally optimize black-box functions. Although Evolutionary Algorithms are quite popular, they require many function evaluations, easily becoming prohibitive when each evaluation is expensive, in terms of time, money, or resources (e.g., raw materials or computational resources). Recent research activities have been pulled by the two drivers which will be both represented in the session. One is the growing set of application domains, including for instance Safe Bayesian optimization, fair and energy aware algorithmic design, and microcontroller design, but also robot navigation and the ensuing issue of stochastic optimal control. The other is the increasing complexity of new methodological and computational challenges. Several issues will be dealt with in the session among which high dimensional problems, combinatorial optimization, noisy and robust optimization, constrained optimization, multi-objective optimization, multi-task, multi-source, and multi-fidelity optimization, distributed and federated BO, BO on manifolds and BO on probabilistic spaces. Effective solutions of these challenges require a deep reformulation of the search space as well as the probabilistic surrogate model and the acquisition function which are at the core of the BO algorithm.

Location, travel, accommodation

Nice, France

LION17 in Nice offers the opportunity to reconnect with colleagues and exchange late-breaking research in a dynamic and modern city. Located on the French Riviera, the southeastern coast of France on the Mediterranean Sea, at the foot of the French Alps, Nice is the seventh most populous urban area in France. Nice is approximately 13 kilometers from the principality of Monaco and 30 kilometers from the French–Italian border ( Nice Wikipedia ).

Nice offers simple travel connections by air, rail and road from all over Europe, and beyond. Nice's airport serves as a gateway to the region and has many flights from the world with affordable cost and decent travel times. From major European cities, Nice can be conveniently reached by plane within 4 to 6 hours. The airport also offers direct flights to worldwide destinations, including to JFK airport in New York City, USA.


The conference takes place at the Hotel Aston La Scala in Nice. The venue is a premier destination located in the heart of Nice, on the French Riviera. Just a short 5-minute walk from the beaches, the venue offers the participants the perfect blend of idyllic setting and modern amenities. The venue is also easily accessible from the airport and is well connected to public transport. While participants can book rooms at the venue (you can use the code IBM to get 10 % off) there are dozens of hotels in close proximity that are also suitable.

Image from:


The 17th Learning and Intelligent Optimization (LION) conference is planned to take place as a physical conference on June 4-8, 2023 in Nice, France . The 2023 LION Organization is closely monitoring the ongoing COVID-19 situation. The safety and well-being of all conference participants is our top priority. After studying and evaluating the announcements, guidance, and news released by relevant national departments, we are prepared to convert LION17 into a hybrid or virtual conference experience. The dates of the conference will remain the same.

Future LION


We are accepting proposals to organize the future editions of LION. If you are interested please signal it to the head of the steering committee (roberto.battiti((AT)) Deadline for proposals: May 31 2023. You may use this template when writing your proposal:


Chair and Local Chairs

Organization committee: Kevin Tierney (Bielefeld University, Germany), Meinolf Sellmann (InsideOpt, USA), Paul Shaw (IBM, France)

Head of steering committee:: Roberto Battiti (University of Trento, Italy)

Technical Program Committee (provisional):

  • Meinolf Sellmann (Chair) (InsideOpt, USA)
  • Kevin Tierney (Co-Chair) (Bielefeld University, Germany)
  • Carlos Ansòtegui (University of Lleida, Spain)
  • Francesco Archetti (Consorzio Milano Ricerche, Italy)
  • Annabella Astorino (ICAR-CNR, Italy)
  • Hendrik Baier (Eindhoven University of Technology, The Netherlands )
  • Roberto Battiti (University of Trento, Italy)
  • Laurens Bliek (Eindhoven University of Technology, The Netherlands )
  • Christian Blum (Spanish National Research Council (CSIC), Spain)
  • Mauro Brunato (University of Trento, Italy)
  • Zaharah Bukhsh (Eindhoven University of Technology, The Netherlands )
  • Sonia Cafieri (Ecole Nationale de l'Aviation Civile, France)
  • Antonio Candelieri (University of Milano Bicocca, Italy)
  • John Chinneck (Carleton University, Canada)
  • Konstantinos Chatzilygeroudis (University of Patras, Greece)
  • Philippe Codognet (JFLI / Sorbonne Universitè, Japan / France)
  • Patrick De Causmaecker (Katholieke Universiteit Leuven, Belgium)
  • Renato De Leone (University of Camerino, Italy)
  • Clarisse Dhaenens (Université Lille 1 (Polytech Lille, CRIStAL, INRIA), France)
  • Luca Di Gaspero (DPIA - University of Udine, Italy)
  • Bistra Dilkina (University of Southern California, USA)
  • Theresa Elbracht (Bielefeld University, Germany)
  • Adil Erzin (Sobolev Institute of Mathematics)
  • Giovanni Fasano (University Ca'Foscari of Venice, Italy)
  • Paola Festa (University of Napoli FEDERICO II, Italy)
  • Adriana Gabor (Khalifa University, Abu Dhabi)
  • Jerome Geyer-Klingeberg (Celones, Germany)
  • Isel Grau (Eindhoven University of Technology, The Netherlands )
  • Vladimir Grishagin (Nizhni Novgorod State University, Russia)
  • Mario Guarracino (ICAR-CNR, Italy)
  • Ioannis Hatzilygeroudis (University of Patras, Greece)
  • Youssef Hamadi (Tempero, France)
  • Andre Hottung (Bielefeld University, Germany)
  • Laetitia Jourdan (INRIA/LIFL/CNRS, France)
  • Serdar Kadioglu (Brown University, USA)
  • Marie-Eleonore Kessaci (Université de Lille, France)
  • Michael Khachay (Krasovsky Institute of Mathematics and Mechanics, Russia)
  • Elias B. Khalil (University of Toronto, Canada)
  • Zeynep Kiziltan (University of Bologna, Italy)
  • Yury Kochetov (Sobolev Institute of Mathematics, Russia)
  • Ilias Kotsireas (Wilfrid Laurier University, Waterloo, Canada)
  • Dmitri Kvasov (DIMES, University of Calabria, Italy)
  • Dario Landa-Silva (University of Nottingham, United Kingdom)
  • Hoai An Le Thi (Université de Lorraine, France)
  • Daniela Lera (University of Cagliari, Italy)
  • Yuri Malitsky (FactSet, USA)
  • Vittorio Maniezzo (University of Bologna, Italy)
  • Silvano Martello (University of Bologna, Italy)
  • Yannis Marinakis (Technical University of Crete, Greece)
  • Nikolaos Matsatsinis (Technical University of Crete, Greece)
  • Laurent Moalic (University of Haute-Alsace - IRIMAS, France)
  • Hossein Moosaei (Jan Evangelista Purkyně University, Czech Republic)
  • Tatsushi Nishi (Osaka University, Japan)
  • Panos Pardalos (University of Florida, USA)
  • Axel Parmentier (Ecole Nationale des Ponts et Chaussées, France)
  • Konstantinos Parsopoulos (University of Ioannina, Greece)
  • Vincenzo Piuri (Universita' degli Studi of Milano, Italy)
  • Oleg Prokopyev (University of Pittsburgh, USA)
  • Helena Ramalhinho (Universitat Pompeu Fabra, Spain)
  • Michael Römer (Bielefeld University, Germany)
  • Massimo Roma (SAPIENZA Universita' of Roma, Italy)
  • Valeria Ruggiero (University of Ferrara, Italy)
  • Frédéric Saubion (University of Angers, France)
  • Andrea Schaerf (University of Udine , Italy)
  • Elias Schede (Bielefeld University, Germany)
  • Marc Schoenauer (INRIA Saclay Île-de-France, France)
  • Marc Sevaux (Lab-STICC, Université de Bretagne-Sud, France)
  • Paul Shaw (IBM, France)
  • Dimitris Simos (SBA Research, Austria)
  • Thomas Stützle (Université Libre de Bruxelles (ULB), Belgium)
  • Tatiana Tchemisova (University of Aveiro, Portugal)
  • Gerardo Toraldo (Università della Campania “Luigi Vanvitelli”, Italy)
  • Paolo Turrini (University of Warwick, UK)
  • Michael Vrahatis (University of Patras, Greece)
  • Om Prakash Vyas (Indian Institute of Information Technology , India)
  • Ranjana Vyas (Indian Institute of Information Technology , India)
  • Dimitri Weiß (Bielefeld University, Germany)
  • Daniel Wetzel (Bielefeld University, Germany)
  • David Winkelmann (Bielefeld University, Germany)
  • Dachuan Xu (Beijing University of Technology, Chine)
  • Qingfu Zhang (University of Essex & City U of HK, Hong Kong)
  • Anatoly Zhigljavsky (Cardiff University, United Kingdom)
  • Antanas Zilinskas (Vilnius University, Lithuania)