Tutorials will be held on July 9, 2017. Traditionally, tutorials attract a broad range of audiences, including professionals, researchers from academia, students, and practitioners, who wish to enhance their knowledge in the selected tutorial topic. Tutorials offer a unique opportunity to disseminate in-depth information on specific topics in fuzzy sets and systems, soft computing/computational intelligence, and related areas.
In the following, the details of tutorials which will be held at FUZZ-IEEE 2017 including the name, the description, the presenters and their biographies (by clicking on the photos) are given.
Fuzzy systems have been used widely thanks to their ability to successfully solve a wide range of problems in different application fields. However, their replication and application requires a high level of knowledge and experience. Furthermore, few researchers publish the software and/or source code associated with their proposals, which is a major obstacle to scientific progress in other disciplines regarding both academy and industry. In recent years, most fuzzy system software (FSS) has been developed in order to facilitate the use of fuzzy systems. Some software is commercially distributed but most software is available as free and open source software, reducing such obstacles and providing many advantages: quicker detection of errors, innovative applications, faster adoption of fuzzy systems, etc. As advised by the Task Force on Fuzzy Systems Software in the Fuzzy Systems Technical Committee of the IEEE Computational Intelligence Society, researchers and developers should think carefully about some critical considerations (interoperability, novelty, usability, and relevance) when publishing a new software.
Moreover, the Standards Committee of the IEEE Computational Intelligence Society has recently sponsored a new standard fuzzy language, namely IEEE Std. 1855TM –2016, which exploits the benefits offered by the eXtensible Markup Language (XML) specifications and related tools in order to make easier modeling fuzzy logic systems in a human-readable and hardware independent way. This tutorial is aimed at bridging the gap in designing fuzzy systems by providing its attendees with a systematic and complete description of state-of-the art of fuzzy system software.
This tutorial is made up of two main parts:
1) Fuzzy Systems Software Overview: Taxonomy, Current Research Trends and Prospects
2) Description of the IEEE Standard for Fuzzy Markup Language
In the first part of this tutorial, J. Alcala-Fdez and Jose M. Alonso (Chair and Vice-chair of the Task Force on Fuzzy Systems Software in the Fuzzy Systems Technical Committee of the IEEE Computational Intelligence Society) will present an overview of freely available and open source FSS in order to provide a well-established framework that helps researchers to find existing proposals easily and to develop well founded future work. They will first introduce a two-level taxonomy and they will briefly discuss some of the main contributions related to each field (Fuzzy Sets Theory, Control, Decision-Making, Software Engineering, etc.). Then, they will provide a snapshot of the status of the publications and citations in this field according to the Thomson Reuters Web of Science, formerly widely known as Institute for Scientific Information (ISI) Web of Knowledge. Then, they will sketch some critical considerations (interoperability, novelty, usability, and relevance) regarding recent trends and potential research directions.
In the second part of the tutorial, G. Acampora and A. Vitiello will introduce the new IEEE Std. 1855TM recently published by IEEE. The Standards Committee of the IEEE Computational Intelligence has published a new standard fuzzy language which exploits the benefits offered by the eXtensible Markup Language (XML) specifications and related tools in order to make easier modeling fuzzy logic systems in a human-readable and hardware independent way. The goal of this part of the tutorial is 1) to show how IEEE 1855 is used to model different kinds of fuzzy systems, and 2) how it can be integrated in different software architecture to enable systems’ designers to embed fuzzy reasoning in their frameworks.
1) J. Alcalá-Fdez and J.M. Alonso, A Survey of Fuzzy Systems Software: Taxonomy, Current Research Trends and Prospects, IEEE Transactions on Fuzzy Systems, 24(1):40-56, 2016, DOI:10.1109/TFUZZ.2015.2426212. Complementary material is available online at http://sci2s.ugr.es/fss
2) IEEE-SA Standards Board, IEEE Std. 1855TM –2016, IEEE Standard for Fuzzy Markup Language, IEEE Computational Intelligence Society, sponsored by the IEEE Standards Committee, 2016.
In the era of the information technology, the problem of managing Big Data applications is becoming the main focus of attention in a wide variety of disciplines such as science, business, industry, and so on. Data and the ability to process and extract knowledge from it are the "new gold" in the digital economy in which we move. As a result, it has emerged an area called Data Science. It collects all scenarios in which data has a starring role with the aim of turning it into knowledge. Data Science encompasses the areas known as machine learning, data mining, social mining, Big Data, and so on.
Addressing Big Data becomes a very interesting and challenging task where we must consider new paradigms to develop scalable algorithms. The MapReduce framework, introduced by Google, allows us to carry out the processing of these large amounts of information. Its open source implementation, named Hadoop, led the development of a popular platform with a wide use. Recently, new alternatives to the standard Hadoop-MapReduce framework have arisen to improve the performance in this scenario, being the most relevant one the Apache Spark project.The MapReduce framework implies that existing algorithms have to be redesigned or that new ones need to be developed in order to take advantage of their capabilities in the big data context.
The challenges posed by real-world Big Data problems are manifold. Apart from the straightforward computational complexity, researchers in this field must deal with vague, imprecise or uncertain data. Among different techniques for the data modeling, those based on fuzzy sets and fuzzy logic are a valuable tool for developing robust solutions. In this tutorial, we will first provide a gentle introduction to the problem of Big Data as well as the presentation of recent technologies (Hadoop ecosystem, Spark). Then, we will dive into the field of Big Data analytics, introducing machine learning libraries such as Mahout and MLlib.
Afterward, we will go across the topic of fuzzy modeling in the Big Data context. We start by introducing the features and design for the most recent approaches in this field. We aim at defining the direction for the design of powerful algorithms based on fuzzy systems, and how the information extracted with these models can be useful for the experts.
The last part of this tutorial will dig into the concept of Data Science. To do so, we will shortly introduce non-classical problems that have acquired much relevance in the last years, such as multi-label problems, transfer learning, multi-view learning, and so on, discussing the usefulness on the use of fuzzy modeling. Additionally, we will focus on the interpretability to raise some questions to reflect the comprehensibility, beyond the dialectical interpretability-accuracy debate, looking for models performing well, understandable and simple. Finally, we analyze the future opportunities that fuzzy modeling may have with respect to Data Science in terms of the knowledge represented by fuzzy and linguistic rules.
Table of contents and estimated duration:
- A gentle introduction to big data
- Big Data Analytics
- Fuzzy Modeling for big data
- Data Science: New scenarios and Non-classical problems
- Fuzzy Models in Data Science - Beyond interpretability vs accuracy trade-off
- Fuzzy Models in Data Science: Opportunities
General type-2 fuzzy sets and systems are paradigms which enable fine-grained capturing, modelling and reasoning with uncertain information. While recent years have seen increasing numbers of applications from control to intelligent agents and environmental management, the perceived complexity of general type-2 fuzzy sets and systems still makes their adoption a daunting and not time-effective proposition to the majority of researchers.
This tutorial is designed to give researchers a practical introduction to general type-2 fuzzy sets and systems. Over three hours, the modular tutorial will address three main aspects of using and working with general type-2 fuzzy sets and systems:
1. Introduction to General Type-2 Fuzzy Sets and Systems The first component of the tutorial will provide attendees with a concise and practice-led overview of general type-2 fuzzy sets and systems, reviewing the motivation behind their definition, their structure in relation to type-1 and interval type-2 fuzzy sets and systems, as well a set of recent applications.
2. Designing General Type-2 Fuzzy Sets and Systems In the second part of the tutorial, two distinct aspects will be discussed. First, attendees will be given a practical introduction to designing their own general type-2 fuzzy system. Using the online browser-based toolkit JuzzyOnline, participants will be guided in the design of a general type-2 fuzzy system, relating their own design to the design of type-1 fuzzy systems at each stage. Second, the design of general type-2 fuzzy sets will be discussed through a presentation of a key set of recently introduced processes to create general type-2 fuzzy sets from data. 3. Coding General Type-2 Fuzzy Sets and Systems
The final part of the tutorial will focus on the programmatic implementation and use of general type-2 fuzzy sets and systems. Currently available software tools and toolkit for general type-2 fuzzy sets and system applications will be briefly reviewed, highlighting usage areas from inference to the computation of measures such as similarity and distance. Finally, interested participants will be supported in the development of a simple general type-2 fuzzy system based on the freely available Juzzy, Phthon and/or R based general type-2 APIs.
Unmanned aerial vehicles (UAVs), commonly known as drones, are commonly used for a number of missions such as traffic surveillance, search and rescue, orchard monitoring, wildlife protection and infrastructure inspection. However, developing almost perfect flight controllers for UAVs for varying working conditions remains a challenging task due to under actuated dynamics, nonlinearities in their models, aerodynamic damping, and internal and external uncertainties. On the other hand, modelling of these complex systems is also tedious, costly and time consuming. Therefore, an advanced (preferably model-free and learning) control approach may be appropriate to improve the control performance and manoeuvrability of UAVs instead of using conventional controllers, e.g., proportional-integral-derivative (PID) controller. As a model-free control technique, type-1 and type-2 fuzzy logic controllers (FLCs) have already been implemented in many industrial control systems. However, they still lack in real-time implementations for UAVs, especially for type-2 FLCs. Therefore, the main aim of this tutorial is to discuss and present real-time implementations of type-1 and type-2 FLCs for controlling UAVs.
An introduction of state-of-art computer vision algorithms used for six degree-of-freedom estimation of UAVs will be presented, and implementation details with FLCs elaborated. Robotic operating system (ROS) will be introduced, and programming FLCs in ROS environment will be shown.
A real quadrotor UAV flight demo related to vision-based autonomous trajectory tracking will be shown indoors with various FLCs in ROS using computer vision algorithms. Online parameter tuning will also be discussed during the demo. Different scenarios will be considered, e.g. performance of different FLCs will be investigated under different noise conditions.
This tutorial will consist of two parts:
Part 1: Theoretical framework for type-1 and type-2 FLCs, vision-based control and ROS
1) We will focus on theoretical basis and definitions of type-1 and type-2 FLCs, introduce state-of-art computer vision algorithms for six degree of freedom pose estimation, various commercial applications of UAVs and discuss their limitations and some reasons why FLCs may be useful.
2) We will present a complete computer vision-based control structure of type-1 and type-2 FLCs for navigating UAVs in ROS environment. The main focus will be implementation of type-1 and type-2 FLCs for control of UAVs, and integration of computer vision algorithms and fuzzy controls.
Part 2: Real-Time implementation using ROS
1) We will demonstrate autonomous navigation of UAVs and discuss different fuzzy controller performances. We will also talk about online tuning of FLCs and its advantages under different noisy working conditions.
2) A sample program will be provided so that attendees can explore real-time implementations of UAV control using different FLCs.
1. Erdal Kayacan and Mojtaba A. Khanesar, “Fuzzy Neural Networks for Real Time Control Applications, Concepts, Modeling and Algorithms for Fast Learning", 1st Edition, Butterworth-Heinemann, Print Book ISBN:9780128026878. (17 Sept 2015)
2. Erdal Kayacan, Reinaldo Maslim. "Type-2 Fuzzy Logic Trajectory Tracking Control of Quadrotor VTOL Aircraft With Elliptic Membership Functions," in IEEE/ASME Transactions on Mechatronics , vol.PP, no.99, pp.1-1 doi: 10.1109/TMECH.2016.2614672
3. Changhong Fu, Andriy Sarabakha, Erdal Kayacan, Christian Wagner, Robert John, Jonathan M. Garibaldi. "A comparative study on the control of quadcopter UAVs by using singleton and non-singleton fuzzy logic controllers," 2016 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE), Vancouver, BC, Canada, 2016, pp. 1023-1030. doi: 10.1109/FUZZ-IEEE.2016.7737800
4. Changhong Fu, Ran Duan, Dogan Kircali, Erdal Kayacan. Onboard Robust Visual Tracking for UAVs Using a Reliable Global-Local Object Model. Sensors, 16(9), 2016. doi:10.3390/s16091406
The goal of this tutorial is to educate and highlight challenges in modern computer vision (CV) and possible directions, tools and novel ideas that the fuzzy community may contribute. Focus will be placed on fuzzy set (FS) theory and some neural networks (NN), in particular deep learning. We will discuss challenges in CV that are recognized by researchers in the areas of low-, mid- and high-level CV. We review a few standard and modern CV approaches, discuss data sets currently used in the CV community, and we present some FS and NN techniques employed in each area, always with an eye towards where soft computing can make the largest impact. We will also provide an open source fuzzy computer vision toolbox (FCVT). At the end of this tutorial, we have an hour of hands on learning and Q&A with the FCVT for synthetic and community benchmark CV datasets. So this tutorial is a combination of theory and application!
Below is a tentative list of topics.
Part I: Introduction to Computer Vision (CV)
- Why study CV?
- What are the different so-called “levels” of CV?
- CV is steeped in probabilistic and neural approaches
- CV heavy communities include PAMI, ICCV, CVPR, ECCV, NIPS, etc.
- CV is a comparison heavy field (methods and data sets)
- Available CV tools: OpenCV, SimpleCV, VLFeat, etc.
- Push to provide open source code: educational and basis for benchmarking
- David Marr: principles of least commitment and graceful degradation
- Where does FS theory fit, e.g., offer the most pay off and/or make the most sense?
Part II: Feature Learning via Deep Learning
- Benefits of learning vs designing features, what is deep learning, and brief survey of deep learning approaches
- Mathematics, architecture and implementation details related to convolutional neural networks (CNNs)
- Existing work and role of FS theory in deep learning
Part III: Data and Information Fusion in CV
- What is data/information fusion and why does CV need it?
- Introduction to fuzzy integrals, extensions and data-driven learning techniques
- Where to fuse in CV? We will focus on signal, spectrum, feature and decision level fusion
Lessons and Q&A: Open Source Fuzzy Computer Vision Toolbox (FCVT)
- Crisp and fuzzy features
- Crisp and fuzzy classification
- Fuzzy integrals and data driven learning algorithms
- Convolutional neural networks
- Fuzzy deep learning
Fuzzy Computing has been applied to various types of research and development in intelligent robotics until now. This tutorial consists of state of the art in intelligent robotics with a focus on fuzzy computing. First, we explain the history on intelligent robotics and the introduction to intelligent robotics including map building, path planning, navigation, and control. The second part of the talk includes how to apply fuzzy computing to two interesting research topics; (1) simultaneous localization and mapping, and (2) multi-robot formation behaviors. Finally, we show several other research topics related with human-robot communication, and discuss the future direction of intelligent robotics.