Prof. Angelov (MEng 1989, PhD 1993, DSc 2015) is a Fellow of the IEEE, of the IET and of the HEA. He is Governor of the International Neural Networks Society (INNS) being his Vice President for two terms till end of 2020. He holds a Personal Chair in Intelligent Systems at Lancaster University, UK. He has authored or co-authored 350+ peer-reviewed publications in leading journals, peer-reviewed conference proceedings, 6 patents, three research monographs (by Wiley, 2012 and Springer, 2002 and 2019) cited over 12300+ times with an h-index of 57. He is the founding Director of LIRA (Lancaster Intelligent, Robotic and Autonomous systems) Research Centre (www.lancaster.ac.uk/lira) which includes over 50 academics across 15 Departments from all Faculties of the University. He has an active research portfolio in the area of computational intelligence and machine learning and internationally recognised results into online and evolving learning and explainable AI. Prof. Angelov leads numerous projects (including several multimillion ones) funded by UK research councils, EU, industry, UK MoD. His research was recognised by 2020 Dennis Gabor Award “for outstanding contributions to engineering applications of neural networks” as well as ‘The Engineer Innovation and Technology 2008 Special Award’ and ‘For outstanding Services’ (2013) by IEEE and INNS. He is also the founding co-Editor-in-Chief of Springer’s journal on Evolving Systems and Associate Editor of several leading international scientific journals, including IEEE Transactions on Fuzzy Systems, of the IEEE Transactions on Cybernetics, IEEE Transactions on AI as well as of several other journals such as Fuzzy Sets and Systems, Soft Computing, etc. He gave over a dozen plenary and key note talks at high profile conferences. Prof. Angelov was General co-Chair of a number of high profile conferences. He was also a member of International Program Committee of 100+ international conferences (primarily IEEE). More details can be found at www.lancs.ac.uk/staff/angelov

Machine Learning (ML) and AI justifiably attract the attention and interest not only of the wider scientific community and industry, but also society and policy makers. However, even the most powerful (in terms of accuracy) algorithms such as deep learning (DL) can give a wrong output, which may be fatal. Due to the hyper-parametric, cumbersome and opaque model used by DL, some authors started to talk about a dystopian “black box” society. Despite the success in this area, the way computers learn is still principally different from the way people acquire new knowledge, recognise objects and make decisions. People do not need a huge amount of annotated data. They learn by example, using similarities to previously acquired prototypes, not by using parametric analytical models. Current ML approaches are focused primarily on accuracy and overlook explainability, the semantic meaning of the internal model representation, reasoning and its link with the problem domain. They also overlook the efforts to collect and label training data and rely on assumptions about the data distribution that are often not satisfied. The ability to detect the unseen and unexpected and start learning this new class/es in real time with no or very little supervision is critically important and is something that no currently existing classifier can offer. The challenge is to fill this gap between high level of accuracy and the semantically meaningful solutions. The most efficient algorithms that have fuelled interest towards ML and AI recently are also computationally very hungry – they require specific hardware accelerators such as GPU, huge amounts of labeled data and time. They produce parameterised models with hundreds of millions of coefficients, which are also impossible to interpret or be manipulated by a human. Once trained, such models are inflexible to new knowledge. They cannot dynamically evolve their internal structure to start recognising new classes. They are good only for what they were originally trained for. They also lack robustness, formal guarantees about their behaviour and explanatory and normative transparency. This makes problematic use of such algorithms in high stake complex problems such as aviation, health, bailing from jail, etc. where the clear rationale for a particular decision is very important and the errors are very costly. All these challenges and identified gaps require a dramatic paradigm shift and a radical new approach.
This Tutorial will introduce step by step this dramatically new approach towards the next generation of computationally lean ML and AI algorithms that can learn in real-time using normal CPUs on computers, laptops, smartphones or even be implemented on chip that will change dramatically the way these new technologies are being applied. It is explainable-by-design. It focuses on addressing the open research challenge of developing highly efficient, accurate ML algorithms and AI models that are transparent, interpretable, explainable and fair by design. Such systems are able to self-learn lifelong, and continuously improve without the need for complete re-training, can start learning from few training data samples, explore the data space, detect and learn from unseen data patterns, collaborate with humans or other such algorithms seamlessly.

The tutorial will provide a round overview picture of the developments in the field of evolving fuzzy systems (EFS) achieved during the last two decades since their first time appearance at the beginning of this century.
Opposed to conventional fuzzy systems, EFS can be learnt from data (streams) on the fly during (fast) on-line processes (typically in form of data streams) in an incremental and mostly single-pass manner. They enjoy a flexible model structure that is able to automatically self-evolve and self-adapt to changes in the process, as e.g. caused by system drifts, new operation modes or dynamic environmental conditions. As being equipped with specific structural components in form of linguistically readable rules, they are able to offer some sort of interpretability and thus to gain insights for operators and experts into system behaviors and dependencies.
They stand for a very important topic in the field of Soft Computing to address modeling problems in nowadays real-world applications with quickly increasing complexity, more and more implying a shift from batch off-line model design phases (as conducted since the 80ties) to permanent on-line (active) model teaching and adaptation cycles toward enriched human-machine interactions. Furthermore, they are widely used in the context of on-line data stream mining and incremental extraction of models and knowledge from huge data bases and Big Data and therefore a fruitful contribution to the research fields of Evolving Adaptive Intelligent Systems and incremental, on-line Machine Learning.
A particular emphasis in this tutorial will be placed on advanced aspects in terms of improved robustness, interpretability and explainability as well as increasing the useability of EFS (Part 2 of the tutorial), comprising issues in the direction of drift handling, incremental curse of dimensionality reduction, dealing with prediction uncertainty, on-line active learning as well as on-line EFS ensemble methods.
Another emphasis will be placed on the application of EFS in the field of quality control and predictive maintenance (Part 3 of the tutorial), which are key applications within the context of “Industry 4.0” and “Factories of the Future (FoF)” Objectives within the current EU framework programmes, to discuss recent advances and to point out still open problems. Thereby, the tutorial speaker will address three variants for on-line monitoring and supervision of industrial (production) systems:

• Retrospective quality control based on user-friendly on-line, adaptive visual inspection systems
• Immediate quality control based on on-line plausibility analysis and condition monitoring of process data
• Predictive quality control (maintenance) based on forecasting models with ideally large prediction horizons (addressing multi-stage processes)

Deep learning has brought the most profound contribution towards living standards by addressing several real-world problems. The continuous advancements in the approaches for classification, localization, segmentation, detection, etc. have extended its application spectrum across various domains. To accomplish such tasks, the models are required to be trained using a huge amount of annotated or labeled data. However, the generation of the annotations for such huge data requires expert analysts and extensive manual efforts. It is a tedious and expensive task, while also being vulnerable to error. Self-supervised learning is an emerging technology that advances to address this issue by effectively closing the gap with fully supervised methods. Here, the aim is to perform pre-training with an unsupervised strategy for learning useful and better representations of the data samples. The pre-trained model is then fine-tuned with limited annotated samples to adopt the actual task such as segmentation, classification, etc. In light of this, the present tutorial discovers various self-supervised learning strategies that could be utilized to improve the performance of the deep learning models.

Deep neural networks (DNNs) have achieved great success in many fields, such as computer vision, natural language processing, speech processing, and so on. The novel architecture is the key factor that decides the performance of DNNs. However, most recent state-of-the-art DNNs are manually designed by experts with rich knowledge, which is not held by each researcher interested. Neural architecture search (NAS) is a research topic that can automatically design high-performance DNN architecture without relying on expert experience. According to different search strategies, NAS can be divided into the Reinforcement learning-based NAS (RL-based NAS), the gradient-based NAS, and the evolutionary computation-based NAS (ENAS). Among the three categories, the ENAS is stand out by its insensitiveness to the local minima and does not require gradient information.

This tutorial will comprehensively introduce ENAS from different perspectives. First, the key steps of designing the ENAS algorithm will be presented, including encoding space, encoding strategy, and population updating. Second, some efficient evaluation strategies for accelerating population updating, which is the key to addressing the efficiency bottleneck of ENAS, are introduced and discussed. Thirdly, recent real-world applications and challenges with potential solutions will be provided to inspire engagement in this field.

The tutorial teaches and demonstrates that the third generation of artificial neural networks, the spiking neural networks (SNN) are not only capable of deep, incremental learning of temporal or spatio-temporal data, but also enabling the extraction of knowledge representation from the learned data and tracing the knowledge evolution over time from the incoming data. Similarly, to how the brain learns, these SNN models do not need to be restricted in number of layers, neurons in each layer, etc. as they adopt self-organising learning principles of the brain. The tutorial consists of the following parts:
1. Why brain-inspired computation?
2. Evolving connectionist systems. EFuNN, DENFIS and NeuCom.
3. Spiking neural networks – a primer.
4. Brain-inspired SNN for deep learning. NeuCube
5. Neuromorphic hardware
6. Design and implementation of SNN systems
7. Applications
8. Conclusions and further directions
The material is illustrated on an exemplar SNN architecture NeuCube (free software and open source along with a cloud-based version available from www.kedri.aut.ac.nz/neucube and www.neucube.io). Case studies are presented including: predictive modelling of EEG and fMRI data measuring cognitive processes and response to treatment; AD prediction; understanding depression; predicting environmental hazards and extreme events.
References:
N.Kasabov, Time-Space, Spiking Neural Networks and Brain-Inspired Artificial Intelligence, Springer, 2019, https://www.springer.com/gp/book/9783662577134
https://theneucom.com     https://www.kedri.aut.ac.nz/neucube

Nikola Kasabov is Professor of Knowledge Engineering and Founding Director KEDRI at Auckland University of Technology (AUT), New Zealand. He holds also George Moore Chair Professor of data analytics at the University of Ulster UK and a Visiting Professorship at the Institute of Information and Communication Technologies at the Bulgarian Academy of Sciences. He is an Honorary Academic at the University of Auckland, the Teesside University and Peking University. Kasabov is a Past President of APNNS and INNS and a leading international scientist in the areas of computational intelligence, neural networks, evolving systems, brain-inspired computation, neuroinformatics, bioinformatics and personalised modelling with a large number of publications and citations. He has received numerous awards. More information can be found from: http://academics.aut.ac.nz/nkasabov.

Ponnuthurai Nagaratnam Suganthan received the B.A degree and M.A degree in Electrical and Information Engineering from the University of Cambridge, UK in 1990, 1992 and 1994, respectively. He received an honorary doctorate (i.e. Doctor Honoris Causa) in 2020 from University of Maribor, Slovenia. After completing his PhD research in 1995, he served as a pre-doctoral Research Assistant in the Dept of Electrical Engineering, University of Sydney in 1995–96 and a lecturer in the Dept of Computer Science and Electrical Engineering, University of Queensland in 1996–99. Since August 2022, he has been with KINDI Centre for Computing Research, Qatar University. He was an Editorial Board Member of the Evolutionary Computation Journal, MIT Press (2013-2018). He is/was an associate editor of the Applied Soft Computing (Elsevier, 2018-), Neurocomputing (Elsevier, 2018-), IEEE Trans on Cybernetics (2012 - 2018), IEEE Trans on Evolutionary Computation (2005 - 2021), Information Sciences (Elsevier) (2009 - ), Pattern Recognition (Elsevier) (2001 - ) and IEEE Trans on SMC: Systems (2020 - ) Journals. He is a founding co-editor-in-chief of Swarm and Evolutionary Computation (2010 - ), an SCI Indexed Elsevier Journal. His co-authored SaDE paper (published in April 2009) won the "IEEE Trans. on Evolutionary Computation outstanding paper award" in 2012. His research interests include randomization-based learning methods, swarm and evolutionary algorithms, pattern recognition, deep learning and applications of swarm, evolutionary & machine learning algorithms. He was selected as one of the highly cited researchers by Thomson Reuters Science Citations yearly from 2015 to 2022 in computer science. He served as the General Chair of the IEEE SSCI 2013. He has been a member of the IEEE (S'91, M'92, SM'00, Fellow 2015) since 1991 and an elected AdCom member of the IEEE Computational Intelligence Society (CIS) in 2014-2016. He is an IEEE CIS distinguished lecturer (DLP) in 2018-2021.

This tutorial will first introduce the main randomization-based feedforward learning paradigms with closed-form solutions. The popular instantiation of the feedforward neural networks is called random vector functional link neural network (RVFL) originated in early 1990s. Other feedforward methods included in the tutorials are random weight neural networks (RWNN), extreme learning machines (ELM), Stochastic Configuration Networks (SCN), Broad Learning Systems (BLS), etc. We will also present deep random vector functional link implementations. Another randomization-based paradigm is the random forest which exhibits competitive performances in batch mode classification. Oblique random forests will also be presented. The tutorial will also consider computational complexity with increasing scale of the classification/forecasting problems. The tutorial will also present extensive benchmarking studies using classification and forecasting datasets.

Job shop scheduling (JSS) has attracted attention from both academics and industry due to their practical value for real-world applications. Evolutionary computation such as genetic programming, has been widely used to find solutions for job shop scheduling problems directly. Instead of heuristic approaches, hyper-heuristics approaches aim to select or generate a heuristic for a class of problems by searching in heuristic spaces. The job shop scheduling problems, genetic programming and evolutionary machine learning techniques investigated in this tutorial are popular investigated problems or techniques in the computational intelligence community.

This tutorial will provide a comprehensive introduction to evolutionary machine learning techniques for JSS. This tutorial will cover different types of (advanced) evolutionary machine learning approaches for JSS. From this tutorial, you are expected to get familiar with evolutionary machine learning in four aspects. First, you will learn the definition of hyper-heuristic learning with a comparison of heuristic learning. Second, the details of JSS (e.g., static, dynamic, flexible JSS) will be given. Third, how to use GP as hyper-heuristic approaches to learn heuristics for JSS will be introduced with examples. Last, this tutorial will show how to use advanced machine learning techniques such as feature selection, surrogate and multitask learning with GP to JSS. All the techniques mentioned will be introduced with promising results.

Artificial intelligence (AI) has entered an increasing number of different domains. A growing number of people – in the general public as well as in research – have started to consider a number of potential ethical challenges and legal issues related to the development and use of AI technologies.
Among the most important challenges are those related to privacy, fairness, transparency, safety and security. Countermeasures can be taken first at design time, second, when a user should decide where and when to apply a system and third, when a system is in use in its environment. In the latter case, there will be a need for the system by itself to perform some ethical reasoning if operating in an autonomous mode.
In this tutorial, we would target to present a structured overview with a focus on the most important ethical issues and their countermeasures. It will also introduce some examples from our own and others ´ work and how the challenges can be addressed both from a technical and human side. AI ethical issues should not be seen only as challenges but also as new research opportunities contributing to more sustainable, socially beneficial services and systems.

Evolutionary multi-objective optimization (EMO) has been a major re- search topic in the field of evolutionary computation for many years. It has been generally accepted that combination of evolutionary algorithms and traditional optimization methods should be a next generation multi-objective optimization solver. As the name suggests, the basic idea of the decomposition-based technique is to transform the original complex problem into simplified subproblem(s) so as to facilitate the optimization. Decomposition methods have been well used and studied in traditional multi-objective optimization. MOEA/D decomposes a multi-objective problem into a number of subtasks, and then solves them in a collaborative manner. MOEA/D provides a very natural bridge between multi-objective evo- lutionary algorithms and traditional decomposition methods. It has been a commonly used evolutionary algorithmic framework in recent years.

Within this tutorial, a comprehensive introduction to MOEA/D will be given and selected research results will be presented in more detail. More specifically, we are going to (i) in- troduce the basic principles of MOEA/D in comparison with other two state-of-the-art EMO frameworks, i.e., Pareto- and indicator-based frameworks; (ii) present a general overview of state-of-the-art MOEA/D variants and their applications; (iii) discuss the future opportunities for possible further developments.

The tutorial will cover emerging security threats of ML Models including algorithm biases, adversarial attacks and the state-of-the-art mitigate strategies. First, we will discuss how a ML systems vulnerability, present different types of adversarial attacks, their severity and applicability of AI/ML-based security. We will discuss limitations of existing defenses in their implementation. Following that, we will suggest possible research directions in addressing adversarial learning challenges: The tutorial topics will be covered (2 hours):
1. AI/ML based systems and the possible vulnerabilities.
2. Adversarial Machine Learning (AML), Algorithm bias, Algorithm verification and validation strategies.
3. Case Studies: AI/ML threats and possible impact on industry.
4. Different computational strategies to mitigate adversarial attacks
5. Challenges and research opportunities to establish safe and secure AI/ML systems.

Dr. Dipankar Dasgupta is a Professor of Computer Science at the University of Memphis; he completed his Ph.D in 1994 in the area of nature-inspired algorithms for Search and Optimization. His research interests are broadly in the area of scientific computing, design, and development of intelligent solutions inspired by biological processes. His book, “Immunological Computation”, is a graduate level textbook, was published by CRC press in 2009. He also edited two books: one on Evolutionary Algorithms in Engineering Applications (1996) and the other is entitled "Artificial Immune Systems and Their Applications", published by Springer-Verlag in 2008 . His latest textbook on Advances in User Authentication is published by Springer-Verlag, 2016.
Dr. Dasgupta has more than 300 publications with 19000+ citations and having h-index of 62 as per Google scholar. He received four Best Paper Awards at international conferences (1996, 2006, 2009, and 2012) and two Best Runner-Up Paper Awards (2013 and 2014). Among other awards, he is the recipient of 2012 Willard R. Sparks Eminent Faculty Award, the highest distinction and most prestigious honor given to a faculty member by the University of Memphis. Prof. Dasgupta received the 2014 ACM SIGEVO Impact Award, an ACM Distinguished Speaker from 2015-2020, and currently IEEE Distinguished Lecturer until 2024.

In the fuzzy system community, data streams are handled by self-adaptive neuro-fuzzy systems (SANFS), which have a single-pass learning mode and an open structure to handle fast, changing data streams. Training SANFSs requires changing numerous free parameters. We proposed a revolutionary SANFS called the parsimonious learning machine (PALM). PALM uses a fuzzy rule based on hyperplane clustering, which reduces the number of network parameters by a factor of 10. Hyperplane-based fuzzy rules are created, merged, and tuned in one pass. rPALM uses teacher-forcing from deep learning literature. PALM's effectiveness was tested using public data and our own autonomous car research. The recommended model reduces processing complexity and the number of critical parameters compared to well-known SANFSs, while maintaining or improving prediction accuracy. Parsimonious Controller (PAC) is a versatile micro aerial vehicle controller (MAVs). It has bias- and variance-based rule-growing and pruning. PAC updates parameters via SMC. LaSalle-Yoshizawa theorem guarantees closed-loop control system boundedness and convergence. Using a bio-inspired flapping wing micro aerial vehicle (BI-FWMAV) and a hexacopter, the controller's performance is assessed.

Advanced data analysis using AI approach plays an important role in Industry 4.0. This talk introduces fundamental concepts of manufacturing data analysis. Several industrial case studies are presented to show how AI has been used for root cause analysis, predictive quality and predictive maintenance in a smart shopfloor. Moreover, Dr Sun Yajuan will share some challenges in industrial application based on her own experiences.