Special sessions

Detail description: 
Real-life problems faced by companies are becoming increasingly complex. As a consequence, hybrid approaches for addressing combinatorial optimization problems are highly popular. The hybridization of metaheuristics with machine learning techniques has been an emerging research field in OR in the last years. From specifically-located hybridizations (parameter fine-tuning, initialization, evaluation, population management, perturbation, local search) to global hybridizations (reduction of the search space, algorithm selection, hyperheuristics, cooperative strategies).

This session is aimed to discuss recent advances and explore future directions in the synergy between metaheuristics and learning mechanisms. The topics of interest include, but are not limited to:

• Learning in nature-inspired computing
• Deep learning in metaheuristics
• Reinforcement learning in metaheuristics
• Performance studies of learning algorithms in metaheuristics
• Feature selection in optimization
• Instance-based metaheuristics
• Learning for operators selection
• Real applications

Description:

Smart cities are based on the synergistic application of communication and information technologies, interconnected devices and information from sensors to improve the quality of life of their citizens.

 

Many real-life problems arise in modern smart cities, including those related to smart transportation systems, smart buildings, smart communications, and smart grid/energy networks. Innovative resolution approaches have been proposed in the literature in recent years, including those dealing with computational intelligence, learning, optimization, and other novel problem-solving strategies.

 

The Special Session on Computational Intelligence for Smart Cities aims at discussing recent advances and exploring future directions on the application of computational methods to solve a wide range of problems arising in smart cities.

 

The topics of interest include, but are not limited to:

• Learning and data science for smart cities • Computational intelligence for smart energy, energy efficiency and sustainability (environmental, social, economic) • Computational intelligence for logistics • Novel resolution approaches for infrastructure, energy and environmental problems • Optimization and management for smart mobility • Computational intelligence in smart homes and Internet of Things • Computational methods to improve governance and citizenship • Computational intelligence in smart healthcare systems • Computational intelligence in tourism and entertainment • Computational intelligence in circular economy • Cyberphysical systems and Internet of Things • Computational intelligence for security, big data, open data, and software


5. Session "Artificial intelligence for health applications", organized by Farah CHEHADE, Institut Charles Delaunay (ICD), Université de Technologie de Troyes, France, farah.chehade@utt.

Artificial intelligence (AI) can help manage and analyze data, process and make decisions using complex algorithms. With high-speed advancement in technology, AI is progressing in daily activities going from SIRI to smart robots or self-driving cars. AI consists of a set of algorithms in a black box that takes collected data as input and yields estimations or predictions of a monitored state. With rapid AI progression, the industries are producing and processing data at a staggering speed.

In healthcare, AI can improve health by helping in diagnosis or prognosis of disease, treatment or prevention of illness, clinical decision support or recovery, well-being or accompanying of patients, etc. AI is of great importance for hospital inpatient care, as well as in home care, for health monitoring. It also enables well-being around patients via forecasting and optimization in healthcare systems.   

This session looks for original contributions addressing health applications using artificial intelligence.

Topics of interest include, but not limited to:

-          Machine learning algorithms

-          Deep learning algorithms

-          Biomedical signal processing

-          Health diagnosis and prognosis

-          Hospital inpatient care

-          Optimization in health system

-          Forecasting for hospital admissions

-          Clustering of patients

-          Scheduling around patients

-          AI in activities of daily living

-          Health monitoring at home

-          Other applications of artificial intelligence in health

 

6.Session “Hyper-Heuristics and their Applications”, organized by: Gabriel Duflo (gabriel.duflo@uni.lu), Emmanuel Kieffer (emmanuel.kieffer@uni.lu) and Grégoire Danoy (gregoire.danoy@uni.lu), University of Luxembourg, Luxembourg

A  wide range of manually-designed heuristics has been developed to tackle optimisation problems. Generally based on common sense rules, their design can be very tedious. Although they provide reasonable execution time, solutions obtained can be far from the optimal solution when considering NP-hard problems.

Efficient heuristics providing strong and robust quality solutions in a reasonable time require in-depth problem knowledge which is a time-consuming and non-trivial task. In addition, manually-designed methods are typically targeted to specific problems and therefore lack generalisation potential. 

Hyper-heuristics have recently been proposed to address these limitations by directly searching in the heuristics' space. Unlike metaheuristics, hyper-heuristics do not then return a solution for a given instance, but a heuristic for a given problem. They have been proposed as a new paradigm in which the mean to get to the best solution is optimised whereas metaheuristics focus only on the solution itself.

While first hyper-heuristics focused on selecting the best heuristics among existing ones, i.e. “heuristics to choose heuristics”, a more recent trend referred to as “heuristic generation” consists in automatically building heuristics from a set of basic components.

Hyper-heuristics have been used in multiple application domains ranging from function combinatorial optimization, vehicle routing, 2D packing and scheduling problems.

This special session aims at providing a forum for researchers and practitioners from academia and industry on recent advances in the field of hyper-heuristics and their applications to theoretical and real-world problems.

 
 Topics of interest include, but are not limited to:
 • Selective hyper-heuristics
 • Generative hyper-heuristics
 • Perturbation-based hyper-heuristics
 • Evolutionary-based hyper-heuristics
 • Deep learning-based hyper-heuristics
 • Reinforcement learning-based hyper-heuristics
 • Multi-objective hyper-heuristics
 • Parallel hyper-heuristics
 • Hyper-heuristics for metaheuristics design
 • Hyper-heuristics on real-world applications