October 30, 2017

Cross-Disciplinary Sessions

CDSS-01 Computational Intelligence for the Automated Design of Machine Learning and Search
CDSS-02 Computational Intelligence methods for Natural Language Processing
CDSS-03 The Role of Computational Intelligence Technologies in Controlling Borders
CDSS-04 Distributed Computational Intelligence (CI) Methods for Data Analytics Over the Internet of Everything
CDSS-05 Computational Intelligence for Bioinformatics and Computational Biology

CDSS-01 Computational Intelligence for the Automated Design of Machine Learning and Search

Organized by Nelishia Pillay (pillayn32@ukzn.ac.za), Rong Qu (rong.Qu@nottingham.ac.uk)

Machine learning and search algorithms play an imperative role in solving real world problems in industry and business sectors. Systems employing these techniques have contributed to many facets of industry including data mining, transportation, health systems, computer vision, computer security, robotics, software engineering and scheduling amongst others. These systems employ one or more techniques such as neural networks, fuzzy logic, evolutionary algorithms, multi-agent approaches and rule-based systems. Implementation of these techniques require a number of design decisions to be made, e.g. what architecture to use, what parameter values to use, and derivation of problem specific operators. It may also be necessary to employ a hybrid system combining techniques to solve a problem which introduces additional decisions such as which techniques to use and how to combine these techniques. This makes the development of computational systems time consuming, requiring many person-hours. Consequently, there have been a number of initiatives to automate these processes using computational intelligence.

There has been a fair amount of research into parameter tuning and control. The field of auto-ML aims to automate the design of machine learning algorithms so as to produce off-the-shelf machine learning techniques. Attempts to automate neural network architecture design has led to the field of neuroevolution. Research in this area has also been directed at inducing fuzzy
functions, rule-based systems and multi-agent architectures. Hyper-heuristics, which were initially aimed at providing generalized solutions to combinatorial optimization problems, are shown to be effective in the automated design of search techniques. Evolutionary algorithms such as genetic programming and genetic algorithms have made a valuable contribution to this
field.

Scope and Topics

The aim of this special session is to examine and promote recent developments in the field and future directions including the challenges and how these can be overcome.
The topics include but are not limited to:

  • Architecture design, e.g. design of neural networks and multi-agent architectures
  • Automated hybridization of intelligent techniques
  • Auto-ML
  • Automatic programming
  • Derivation of constructive heuristics
  • Derivation of evaluation functions
  • Derivation of operators
  • Explainable machine learning
  • Hyper-heuristics
  • Neuroevolution
  • Parameter control and tuning
  • Search-based software engineering
  • Self*-search

 

CDSS-02 Computational Intelligence methods for Natural Language Processing

Organized by Keeley Crockett (K.Crockett@mmu.ac.uk), Joao Paulo Carvalho (joao.carvalho@inesc-id.pt)

Although language, or linguistic expressions, undoubtedly contains fuzziness in nature, very little research has been conducted in related fields in recent years, as it was shown in “A Critical Survey on the use of Fuzzy Sets in Speech and Natural Language Processing”, Proc. of the IEEE WCCI 2012, Brisbane, Australia. This is partly because of the prevalence of probabilistic machine learning technologies in the natural language processing field. However, there has been a growing recognition that fuzziness found in every aspect of human language has to be adequately captured and that recent developments in the fields of computational intelligence such as computing with words can make a contribution. This session will follow on from the successful, special sessions entitled “Fuzzy Natural Language Processing” which were successfully held at IEEE FUZZ 2017 in Naples, IEEE FUZZ 2015 in Istanbul and IEEE FUZZ 2013 in India and the hybrid special sessions held at both the 2014 IEEE World Congress on Computational Intelligence in Beijing and the 2016 IEEE World Congress on Computational Intelligence in Vancouver.

The aim of this Special Session is therefore to explore new techniques and applications in the field of
computational intelligence approaches to natural language processing.

Scope and Topics

The session will provide a forum to disseminate and discuss recent and significant research efforts in fuzzy, neural and evolutionary methods for natural language processing in addition to hybrid and emerging computational intelligence paradigms. It also seeks to present novel applications of computational intelligence technologies within the field of natural language processing It invites researchers from different related fields and gathers the most recent studies including but not limited to:

  • Fuzzy set models of human language
  • Computational intelligence applications to human language processing
  • Machine learning approaches to human language
  • Computational intelligence approaches to text mining
  • Computational Intelligence simulations of language use
  • Fuzzy ontologies for human language
  • Computational intelligence applications to the semantic web
  • Computing with words within natural language processing
  • Real world computational intelligence inspired natural language processing applications
  • Computational intelligence founded methodologies, tools and techniques for mining and interpretation of social media textual data

 

CDSS-03 The Role of Computational Intelligence Technologies in Controlling Borders

Organized by Keeley Crockett (K.Crockett@mmu.ac.uk), Rodoula Makri, George Boultadakis

Continuous border traffic growth, combined with the increased threat of illegal immigration, is putting border agencies under considerable pressure internationally. Border control is likely to face increasing demands for performance efficiency whilst maintaining high levels of security and conformity to legal frameworks, implying the need for intelligent systems that are user friendly and reliable in operational conditions, overcoming the limitations and potential gaps of current operational procedures. Key challenges are in the design of such systems which harness computational intelligence algorithms whilst allowing human empowerment, through the use of technologies which are familiar to all stakeholders.

To address these challenges multi-disciplinary research needs to be carried out in order to develop comprehensive systems to be designed and implemented which can provide automated computationally intelligence platforms. Recent research examples focused on the land border control include disciplines in the areas of: analysis of the traveller’s non-verbal behaviour, analytics of document authenticity, discovery of key patterns through data-mining and machine learning for border control analytics, hidden human detection to confront illegal immigration, advanced algorithms, big data, artificial intelligence and neural networks along with face, fingerprints and palm vein biometric models. These are just some key examples of how scientific disciplines can be combined together to enable automatic risk assessment enabling reliable decision making at border control points which respect an individual’s privacy whilst maintaining data security.
Based on these emerging research trends, the aims of this Special Session are to:

  • to provide a forum for new computational intelligence methodologies / techniques and systems which contribute towards improving border crossing efficiency and security within border control solutions;
  • to provide the opportunity to present recent advances towards holistic systems that combine computational intelligence within an expanded multi-disciplinary context
  • to highlight and assess novel insights and intelligence that can effectively contribute in identifying threats, vulnerabilities and risks in border control improving decision making and efficiency
  • to investigate the social and ethical implications of using computational intelligence technology on the ‘passenger’ themselves.

Scope and Topics

The session will provide a forum for both academia and industry to disseminate and discuss recent and significant research efforts in fuzzy, neural and evolutionary methods and their integration into border security solutions, in addition to hybrid and emerging computational intelligence paradigms. It also seeks to present novel applications of computational intelligence technologies which contribute to human decision making within the border control lifecycle. Technical papers addressing research challenges in these areas are welcomed.
It invites researchers from different related fields and gathers the most recent studies including but not limited to:

  • Biometric identification using computational intelligence
  • Automated interviews of passengers using intelligent systems
  • Detection of fraudulent passenger documents
  • The role of computational intelligence in advanced passenger registration
  • Interaction of sensors, detection tools and advanced signal processing with intelligent systems
  • Confronting illegal immigration through intelligent systems and detection of hidden humans
  • Automated decision making systems at borders
  • Passenger Identity control and status prediction
  • Intelligent Border Control Analytics
  • Ethical and social impact on the use of computational intelligence based systems with stakeholders

 

CDSS-04 Distributed Computational Intelligence (CI) Methods for Data Analytics Over the Internet of Everything

Organized by Farookh Khadeer Hussain, Mukesh Prasad, Albert Zomaya

Computational Intelligence (CI) is a collection of techniques and methods that have been proven to be capable of solving dynamic, complex, real-world business problems and aiding decision making. The various CI methods have been around for a number of decades and have been used by enterprise sectors across the globe for reliable and accurate decision making. Computational Intelligence (CI) methods include (but are not limited to) neural networks, fuzzy systems, and evolutionary computation.

A critical issue facing the CI world is how to handle, analyse, make sense of and derive critical business insights fromthe large amount of enterprise data that is continuously beinggenerated, particularly since this data originates from devices/nodes that are distributed globally and are not confinedto a particular location. For example, the Internet of Things (IoT) has revolutionized the manner in which enterprise data is created and managed, especially in respect of the volume, velocity,and heterogeneity of enterprise data creation and management. It is conservatively estimated that by 2050 there will be 50 billion IoT nodes, roughly seven times the world’s human population, which leads to the pragmatic notion of an ‘Internet of Everything’ (IoE) in which almost every device will be connected to the Internet and churning out data at an unprecedented rate. Enterprise Cloud platforms have come to the rescue in providing a reliable mechanism for the storage and processing of this data. ‘Cloud-enabled Data Analytics’ or ‘Cloud-based Data Analytics’ is a new and upcoming area which requires the development and use of novel, efficient, optimized and distributed CIalgorithms on top of both the incoming data streams (from IoT nodes) and the static data stored on Cloud platforms.From a Computational Intelligence perspective, the transition from traditional ‘Computational Intelligence’ to ‘Distributed ComputationalIntelligence’ is not automatic and needs to be researched. CI methods need to be developed which will enable real-time decision making on data from distributed sources. The accuracy and speed of these distributed CI algorithms are critical for real-time and reliable decision making. Currently, these are open research issues and challenges which impede the deployment of Distributed Computational Intelligence on top of the ‘Internet of Everything’.

For the ‘Internet of Everything’ to become a reality, there is additionally an urgent need to address the underlying issues of intelligent data storage and data processing at two key storage and processing levels – the Edge Level and the Cloud Level. The Edge level is an intermediate transitory level for data storage; however, it can also be used for efficient data analytics (at the Edge level). There is also a need to engineer new distributed and intelligent CI solutions that will work in tandemwith the Edge level and the Cloudlevel, and will synchronize with each other on either a real-time basis or a quasi-real-time basis depending on the business application at hand.

Scope and Topics

The goal of this special session is to target topics related to distributed, computational intelligence algorithms, methods and techniques for reliable business intelligence on top of the Internet of Everything.We invite interested authors to submit their original and unpublished work to this special session.

The main topics of this special session are as follows:

  • Data Analytics middleware for Edge computing
  • Cloud-based intelligent analytics
  • Edge-node-driven data analytics
  • Intelligent data synchronization and updating between Edge nodes or Cloud nodes
  • Data metering for Edge nodes and Cloud nodes
  • Intelligent pricing mechanisms for Edge nodes and Cloud nodes
  • Data-driven privacy and security solutions in Edge computing and Cloud computing
  • Case studies for data analytics using Edge nodes or Cloud nodes

 

CDSS-05 Computational Intelligence for Bioinformatics and Computational Biology

Organized by Antonello Rizzi (antonello.rizzi@uniroma1.it), Alessandro Giuliani (alessandro.giuliani@iss.it), Alessio Martino (alessio.martino@uniroma1.it)

Bioinformatics and Computational Biology deal with a wide range of problems and applications which, in recent years, have been successfully solved by means of Computational Intelligence and Machine Learning techniques. Moreover, due to technological progress, huge amounts of data concerning biological organisms are gathered and collected (e.g. genes transcript, protein structures and the like), thereby demanding the use of parallel and distributed computing for facing Big Datasets and/or high-throughput application requirements. Further, in such fields, data usually encodes complex information, which is natively represented by structured records, such as sequences, graphs and images, most of which lie in so-called “non-metric spaces”, i.e. input spaces for which a meaningful (dis)similarity measure might not be metric, making the problem more challenging since ad-hoc (dis)similarity measures or embedding functions need to be defined.

Scope and Topics

This Special Session aims at collecting the latest research in Computational Intelligence applications for Bioinformatics and Computational Biology, with emphasis on parallel/distributed computing and non-metric spaces analysis, by means of different (or hybridization of) Computational Intelligence techniques, from evolutionary meta-heuristics to neural computation, from pattern recognition to fuzzy systems.

Topics of interest include (but are not limited to):

  • Protein function prediction
  • Protein folding prediction
  • Generative models for protein contact networks
  • String kernel methods for sequence classification
  • Mining metabolic pathways
  • Gene finding and prediction
  • Exact/inexact motifs and pattern matching
  • Network and Systems Biology
  • Granular computing approaches for non-metric spaces analysis
  • Large-scale data mining and pattern recognition
  • Distributed and parallel computing systems for machine learning and data mining
  • Clinical Diagnostic Systems
  • Medical image analysis