Home
Technical program
Invited and Tutorials
Registration
Venue, Hotels and Social
Paper Submission
Special Sessions
Program Committee
 

LION 12 Call for Papers: Special Sessions

Special sessions are organized as part of LION12 as a way to focus submissions and encourage more interaction between specific communities. In general, submission and publication rules are the same as for the general call for papers, with the organizers of the special sessions coordinating and helping in identifying competent reviewers.

This webpage will be updated as more sessions are proposed and accepted. In case you are interested in organizing a session during LION 12, please feel free to contact the General Chairs.


  • How machine learning is revolutionising healthcare

    Organizers:

    • Dr. Kostas Chrisagis <kostasnc@gmail.com>, City University London, United Kingdom
    • Dr. Serafeim Moustakidis <smoustakidis@gmail.com>, Center for Research and Technology Hellas

    Description:

    The proliferation of massive and heterogeneous health-related data brings with it a series of special challenges enabling at the same time opportunities for improving healthcare. Clinicians and health experts are overwhelmed by the volume, velocity and variety of the available data including medical imagery, data from wearable sensors, electronic health records, genomic data, behavioral and environmental data. The increased availability of data and computational power has led to a resurgence of machine learning leading the efforts to transform the vast amount of complex health-related data into actionable knowledge. Machine learning and deep learning are now attempting to revolutionize the whole healthcare sector by improving diagnostics, predicting outcomes, and changing the way doctors think about providing care. Reflecting this excitement, this special session aims to identify opportunities and challenges of the growing intersection of machine learning and health.

    Topics of interest include but are not limited to:

    • Imaging related decision making and computer-aided diagnosis
    • Multi-modal Clinical Decision Support
    • Machine learning / deep learning for medical image analysis
    • Early detection and diagnosis of diseases
    • Big data analytics in healthcare
    • Data mining with interpretable models
    • Enhanced imaging diagnostics
    • Behavioral analysis with wearables
    • Variable selection over high dimensional heath related data
    • Personalized diagnosis and treatment
    • Drug Discovery using unsupervised learning
    • Computational Methods in Molecular Biology


  • Computational Intelligence for Smart Cities

    Organizers:

    • Enrique​ Alba​ (Professor),​ University​ of​ Malaga,​ Spain (link)
    • Konstantinos​ Parsopoulos​​ (Associate​ Professor),​ University​ of​ Ioannina,​ Greece (link)

    Description:

    Global urbanization is continuously reshaping our world. More than half of the world's population is currently living in urban areas, with predictions adding 2.5 billion people to the cities over the next few decades. This transformation provides great opportunities for cultural and economic growth. However, it also comes along with a number of challenging problems such as overpopulation in metropolitan areas, cost of living, environmental pollution, and inadequate infrastructures, among others. Smart cities attempt to provide solutions to the continuously growing needs by integrating information technologies and interconnected devices in urban environments. This allows the collection and interpretation of huge amounts of data that are used for optimizing various aspects of the cities operation through the design and modeling of ad hoc solutions and systems. Smart transportation systems, smart buildings, smart communications and energy networks are some of the​ ​most​ active​ research​ areas​ in​ this​ domain.

    Computational intelligence (C.I.) has played a significant role in most complex systems existing till now, and they are also expected to have a prominent position in smart cities. Its constituent methodologies such as machine learning, data science, artificial neural networks, evolutionary algorithms, swarm intelligence, and fuzzy logic offer computationally efficient methodologies for modeling, analyzing, and optimizing smart cities systems. Indeed, computational intelligence is one important way to build to the “smart” part of the city. The interplay of such approaches with operations research and many other domains (civil engineering, urban planning, policy makers, companies...) can offer innovative and sustainable solutions to problems of high complexity. So far, computational intelligence methodologies reckons a large number of applications in smart cities, including smart transportation systems, smart logistics, smart energy grids, smart resources integration​ and​ pollution​ monitoring.

    The present special session welcomes works on any aspect of computational intelligence in smart cities​ environments,​ both​ theoretical​ and​ applied,​ including:

    • Computational​ intelligence​ in​ smart​ transportations​ and​ logistics
    • Computational​ intelligence​ in​ urban​ mobility​ and​ planning
    • Computational​ intelligence​ in​ smart​ energy​ systems
    • Computational​ intelligence​ in​ sustainability​ (environmental,​ social,​ economic)
    • Computational​ intelligence​ in​ smart​ homes​ and​ Internet​ of​ Things
    • Computational​ intelligence​ in​ smart​ healthcare​ systems
    • Computational​ intelligence​ in​ governance
    • Computational​ intelligence​ for​ people​ and​ good​ living
    • Computational​ intelligence​ to​ tourism​ and​ entertainment​ in​ the​ city
    • Computational​ intelligence​ in​ circular​ economy
    • Cyberphysical​ systems​ and​ Internet​ of​ Things​ coupled​ with​ C.I.
    • Computational​ intelligence​ for​ security,​ big​ data,​ open​ data,​ and​ software​ for​ cities

    Applications involving efficient learning and optimization methodologies for this type of problems are​ strongly​ encouraged.

    Important​ dates:

    February​ 15,​ 2018​: Paper​ submission
    February​ 28,​ 2018​​: Author​ Notification
    March​ 15,​ 2018​​: Camera​ ready
    June​ 10-15,​ 2018​​: Conference
    July 15, 2018: Submission of camera-ready papers, formatted according to Springer's LNCS guidelines


  • On the borderline between Data Analysis and Combinatorial Optimisation: models, algorithms, and bounds

    Organizers:

    • Prof. Alexander Kelmanov, Sobolev Institute of Mathematics, Novosibirsk, Russia
    • Prof. Michael Khachay, Krasovsky Institute of Mathematics and Mechanics, Ekaterinburg, Russia

    Description:

    Combinatorial optimization and data analysis appear to be extremely close fields of the modern computer science. For instance, various areas in machine learning: PAC-learning, boosting, cluster analysis, feature and model selection, etc. are continuously presenting new challenges for designers of optimization methods due to the steadily increasing demands on accuracy, efficiency, space and time complexity and so on. In many cases, learning problem can be successfully reduced to the appropriate combinatorial optimization problem: max-cut, k-means, p-median, TSP, and so on. To this end, all the results obtained for the latter problem (approximation algorithms, polynomial-time approximation schemes, approximation thresholds) can find their application in design the high-precision and efficient learning algorithms for the former one. On the other hand, there are known examples, where combinatorial optimisation and computational geometry benefits from using approaches developed in statistical learning theory. Among them are Chernoff like measure concentration theorems employed for designing of randomised algorithms and schemas and Bronnimann-Goodrich epsilon-net approach to approximation the famous Hitting Set problem. This session welcomes papers presenting new results on computational and parametric complexity, design and implementation of efficient algorithms and schemes for various extremal problems coming from combinatorial optimisation, classification, clustering, computational geometry, and so on.

    Topics of interest include but are not limited to:

    • computational and parametric complexity
    • inapproximability issues and approximation thresholds
    • polynomial time solvable subclasses of intractable problems
    • polynomial time approximation algorithms and schemes
    • randomized approximation and asymptotically optimal algorithms
    • efficient approximation algorithms for geometric settings of NP-hard problems
    • efficient techniques of supervised, semi-supervised, and unsupervised learning

    Important​ dates:

    February​ 15,​ 2018​: Paper​ submission
    February​ 28,​ 2018​​: Author​ Notification
    March​ 15,​ 2018​​: Camera​ ready
    June​ 10-15,​ 2018​​: Conference
    July 15, 2018: Submission of camera-ready papers, formatted according to Springer's LNCS guidelines


  • Graphical model selection and applications

    Organizers:

    • Dr. Valeriy Kalyagin, Laboratory of Algorithms and Technologies for Network Analysis, National Research University Higher School of Economics, Nizhny Novgorod, Russia
    • Dr. Mario Guarracino, High Performance Computing and Networking Institute, Italian National Research Council, Naples, Italy

    Description:

    Graphical models provide a unifying framework for capturing dependencies in complex systems. Graphical models are recognized as a useful tool in many applied fields, such as bioinformatics, communication theory, combinatorial optimization, signal and image processing, information retrieval, stock market network analysis and statistical machine learning. Graphical model selection is a practical problem of identification of the underlying graphical model from observations. The session will be devoted to theoretical aspects and practical algorithms for graphical model selection and its applications. Estimating the graph structure given a set of observations at the nodes is very common in many fields and in particular in biology, where the complexity of processes and functions are widely modeled by networks. From protein interaction to metabolic pathways, from gene regulatory circuits to brain connectomes, networks have sizes that range from few thousands to many trillions vertices. From their analysis, we can obtain more insights in complex questions, identifying for example their critical points, robustness and modularity. In this session, we will address some of the recent advances on graphical model selection, that can find application in different disciplines and applications.

    Topics of interest include but are not limited to:

    • Graphical model selection in bioinformatics
    • Graphical model selection in communication
    • Graphical model selection in combinatorial optimization
    • Graphical model selection in signal and image processing
    • Graphical model selection in information retrieval
    • Graphical model selection in market network analysis
    • Graphical model selection in statistical machine learning
    • Graphical model selection in gene expression network
    • Graphical model selection in gene co expression network

    Important​ dates:

    February​ 15,​ 2018​: Paper​ submission
    February​ 28,​ 2018​​: Author​ Notification
    March​ 15,​ 2018​​: Camera​ ready
    June​ 10-15,​ 2018​​: Conference
    July 15, 2018: Submission of camera-ready papers, formatted according to Springer's LNCS guidelines


  • Optimization and Management in Smart Manufacturing

    Organizers:

    • Dr. Panos M. Pardalos, <pardalos@ufl.edu>, University of Florida, USA
    • Dr. Xinbao Liu, <lxb@hfut.edu.cn>, School of Management, Hefei University of Technolog, China
    • Dr. Jun Pei, <peijun@hfut.edu.cn>, School of Management, Hefei University of Technolog, China

    Description:

    Current global science and technology innovation shows some new development trends and characteristics. Emerging information technology such as internet and big data technology are widely permeated which drives the group technology revolution characterized by green, intelligent, and ubiquitous in almost all the areas. Intelligentization, greenization, servitization, and interconnection are becoming important for the development and revolution of manufacturing industry. Interdisciplinary and networked innovative platform is reforming the innovation system of traditional manufacturing industry. A new green manufacturing system based on the value network is forming. This session aims to apply data-driven resource management and optimization technique to offer theory support for the management revolution, business pattern revolution, decision theories and methods innovation, and intelligent decision system construction.

    Topics of interest include but are not limited to:

    • Network Manufacturing
    • Sustainability Manufacturing Strategy Management
    • Sustainability Supply Chain Operations Management
    • R&D project management of high-end equipment
    • Environmental and Sustainability Assessment
    • Operations Management of Smart Factory
    • Behavioral Operations Management
    • Production engineering management of high-end equipment
    • Inventory Planning and Control of Green Products
    • Green Logistics Operation and Management
    • Service engineering management of high-end equipment
    • Remanufacturing engineering management of high-end equipment
    • Quality Management Based on Industrial Big Data
    • Development Management of Renewable Energy Technologies
    • Big Data Applications in Smart Manufacturing


  • Algorithms and Applied Optimization for Environmental Data Science (link)

    Organizers:

    • Francesco Archetti, Professor of Computer Science, University of Milano Bicocca
    • Ciprian Dobre, University Politehnica of Bucharest, Computer Science Department

    Description:

    The impact of information technology and the data deluge today led to the appearing of a fourth paradigm of science, namely Data Science. Nowhere this direction led to more impact than in case of Environmental studies. In face of modern challenges faced by our planet, to better understand root causes and research means to help our sustainable future, Environmental scientists in particular rely today more and more on factual data. As such, more environmental data are available now than ever before. Sensors are used to understand phenomena as complex as the Global Warming, to help wildlife better deal with human-induced conditions, to improve crops and farming, and even to deal with causes of pollution in case of cities of tomorrow (in the face of an increasingly more dense population moving to cities worldwide, information technology can automate many of the municipality policies, leading to what today are known as Smart Cities).

    Except for data collected by sensors, environmental data come in many shapes and sizes, like public disclosures, compliance reporting, results from past investigations, internal organizational data and many others. While careful thought usually goes into collection of such data, the workshop is a venue for experts and specialists interested in finding out answers on how to properly analyze the data once it has been generated. Such data can, in fact, generate true information, but for this we need to research innovative approaches, classical statistics, and modern analytical tools. The AODS workshop invites prospective authors to disseminate original contributions on algorithms, machine learning and intelligent optimization, thus the tools to answer questions such as:

    • What are the intended uses of the data set?
    • What else can you learn from the data?
    • How do you optimally manage the data?
    • How do you understand and use the data most effectively?
    • Who needs to use these data? Who else could use it and how?
    • How defensible and legally admissible are the data?

    AODS will thus focus on the following research issues around the tools and ideas in the Environmental Data Scientist’s toolbox, including but not limited to:
    • Mechanisms for data clearing and make data tidy (to speed downstream data analysis tasks), by optimizing the process of data gathering.
    • Exploratory data analysis, to support the development of complex statistical models, or to eliminate or sharpen potential hypotheses about the world that can be addressed by the data.
    • Concepts and tools behind reporting modern data analyses in a reproducible manner.
    • Statistical inference tools, mechanisms and methods, including statistical modelling, data oriented strategies and use of designs and randomization in analyses, and others.
    • Linear models, including model selection and uses of models.
    • Prediction and machine learning, including building and applying prediction functions with an emphasis on practical applications.
    • Data products to automate complex analysis tasks or use technology to expand the utility of a data informed model, algorithm or inference.
    • Practical approaches to deal with data protection and data access models.

    See the detailed description.


  • Machine Learning and Robust Optimization Techniques in Financial Applications

    Organizers:

    • Prof. Theodore B. Trafalis, School of Industrial and Systems Engineering, The University of Oklahoma, USA (Email: ttrafalis@ou.edu)

    Description:

    Traditional financial engineering has focused on stochastic optimization models and their mathematical analysis. Since there are no fundamental laws in financial engineering several restrictive assumptions must be made. A data-driven approach, such as machine learning provides new opportunities for the development of new machine learning models for financial engineering. Today, stocks are frequently traded via electronic exchanges (high frequency trading). Market events are often reported at the nanosecond granularity. In high frequency trading the limit order book data generated over time is of the size of terabytes to petabytes. Therefore new machine learning algorithms are needed to handle these data.

    This session welcomes papers presenting new results on machine learning, applications in financial engineering and energy systems.

    Topics of interest include but are not limited to:

    • Machine Learning in Financial Applications
    • Mathematical Analysis in Machine Learning and Deep Learning
    • Development of new learning models and methods for financial applications
    • Applications of Machine Learning in Financial Engineering
    • Machine Learning for High Frequency Trading
    • Robust Optimization and Learning
    • Machine learning in Energy trading

    Important dates:

    February 22, 2018: Paper submission
    February 28, 2018: Author Notification
    March 15, 2018: Camera ready
    June 10-15, 2018: Conference
    July 15, 2018: Submission of camera-ready papers, formatted according to Springer's LNCS guidelines