Speakers

Gianluca Bontempi

Gianluca Bontempi is Full Professor in the Computer Science Department at the Université Libre de Bruxelles (ULB), Brussels, Belgium, founder and co-head of the ULB Machine Learning Group. His main research interests are big data mining, machine learning, bioinformatics, causal inference, predictive modelling and their application to complex tasks in engineering (forecasting, fraud detection) and life science. He was Marie Curie fellow researcher, he was awarded in two international data analysis competitions and he took part to many research projects in collaboration with universities and private companies all over Europe. In 2013-17 he has been the Director of the ULB/VUB Interuniversity Institute of Bioinformatics in Brussels. He is author of more than 200 scientific publications, member of the scientific advisory board of Chist-ERA and IEEE Senior Member. He is also co-author of several open-source software packages for bioinformatics, data mining and prediction.

"Machine Learning for Predicting in a Big Data World"

The increasing availability of massive amounts of data and the need of performing accurate forecasting of future behavior in several scientific and applied domains demands the definition of robust and efficient techniques able to infer from observations the stochastic dependency between past and future. The forecasting domain has been influenced, from the 1960s on, by linear statistical methods such as ARIMA models. More recently, machine learning models have drawn attention and have established themselves as serious contenders to classical statistical models in the forecasting community. 

See more about Prof. Bontempi here.

Usama Fayyad

CEO Open Insights / www.open-insights.com

He reactivated Open Insights, after leaving Barclays in London. He is also Interim CTO for Stella.AI a Mountain View, CA VC-Funded startup in AI for recruitment. He is acting as Chief Operations & Technology Officer for MTN’s new division: MTN2.0 aiming to extend Africa’s largest telco into new revenue streams beyond Voice & Data.

See more about Mr. Fayyad here.

Peter Flach

Professor of Artificial Intelligence
Intelligent Systems Laboratory, Department of Computer Science 

University of Bristol, UK

"The value of evaluation: towards trustworthy machine learning"

Machine learning, broadly defined as data-driven technology to enhance human decision making, is already in widespread use and will soon be ubiquitous and indispensable in all areas of human endeavour. Data is collected routinely in all areas of significant societal relevance including law, policy, national security, education and healthcare, and machine learning informs decision making by detecting patterns in the data. Achieving transparency, robustness and trustworthiness of these machine learning applications is hence of paramount importance, and evaluation procedures and metrics play a key role in this.

In this talk I will review current issues in theory and practice of evaluating predictive machine learning models. Many issues arise from a limited appreciation of the importance of the scale on which metrics are expressed. I will discuss why it is OK to use the arithmetic average for aggregating accuracies achieved over different test sets but not for aggregating F-scores. I will also discuss why it is OK to use logistic scaling to calibrate the scores of a support vector machine but not to calibrate naive Bayes. More generally, I will discuss the need for a dedicated measurement theory for machine learning that would use latent-variable models such as item-response theory from psychometrics in order to estimate latent skills and capabilities from observable traits.

See more about Prof. Flach here.

Francisco Herrera

Head of Research Group SCI²S
Soft Computing and Intelligent Information Systems

Universidad de Granada, Spain

"A tour on Imbalanced big data classification and applications"

Big Data applications are emerging during the last years, and researchers from many disciplines are aware of the high advantages related to the knowledge extraction from this type of problem.

The topic of imbalanced classification has gathered a wide attention of researchers during the last several years. It occurs when the classes represented in a problem show a skewed distribution, i.e., there is a minority (or positive) class, and a majority (or negative) one. This case study may be due to rarity of occurrence of a given concept, or even because of some restrictions during the gathering of data for a particular class. In this sense, class imbalance is ubiquitous and prevalent in several applications. The emergence of Big Data brings new problems and challenges for the class imbalance problem.

In this lecture we focus on learning from imbalanced data problems in the context of Big Data, especially when faced with the challenge of Volume. We will analyze the strengths and weaknesses of various MapReduce-based algorithms that address imbalanced data. We will present the current approaches presenting real cases of study and applications, and some research challenges.

See more about Prof. Herrera here. 

Enrique Herrera-Viedma

Enrique Herrera-Viedma  is   Professor in Computer Science and A.I in University of Granada and currently the new Vice-President for Research and Knowlegde Transfer. His current research interests include group decision making, consensus models, linguistic modeling, aggregation of information,  information retrieval, bibliometrics, digital libraries, web quality evaluation,  recommender systems, and social media. In these topics he has published more than 200 papers in ISI journals and coordinated more than 20 research projects. Dr. Herrera-Viedma is member of the gobernment board of the IEEE SMC Society and an Associate Editor of  international journals such  as the IEEE Trans. On Syst. Man, and Cyb.: Systems, Knowledge Based Systems, Soft Computing, Fuzzy Optimization and Decision Making, Applied Soft Computing, Journal of Intelligent and Fuzzy Systems, and Information Sciences. 

"Bibliometric Tools for Discovering Information in Science"

In bibliometrics, there are two main procedures to explore a research field: performance analysis and science mapping.Performance analysis aims at evaluating groups of scientific actors (countries, universities, departments, researchers) and the impact of their activity on the basis of bibliographic data. Science mapping aims at displaying the structural and dynamic aspects of scientific research, delimiting a research field, and quantifying and visualizing the detected subfields by means of co-word analysis or documents co-citation analysis. In this talk we present two bibliometric tools that we have developed in our research laboratory SECABA: H-Classics to develop performance analysis by based on Highly Cited Papers and SciMAT to develop science mapping guided by performance bibliometric indicators.

See more information about Prof. Herrera-Viedma here.

Witold Pedrycz

Professor and Canada Research Chair IEEE Fellow
Professional Engineer Department of Electrical and Computer, University of Alberta, Canada

"Linkage Discovery: Bidirectional and Multidirectional Associative MemoriesIn Data Analysis" 

Associative memories are representative examples of associative structures, which have been studied intensively in the literature and have resulted in a plethora of applications in areas of control, classification, and data analysis. The underlying idea is to realize associative mapping so that the recall processes (both one-directional and bidirectional) are characterized by a minimal recall error.

We carefully revisit and augment the concept of associative memories by proposing some new design directions. We focus on the essence of structural dependencies in the data and make the corresponding associative mappings spanned over a related collection of landmarks (prototypes). We show that a construction of such landmarks is supported by mechanisms of collaborative fuzzy clustering. A logic-based characterization of the developed associations established in the framework of relational computing is discussed as well.

Structural augmentations of the discussed architectures to multisource and multi-directional memories involving associative mappings among various data spaces are proposed and their design is discussed.

Furthermore we generalize associative mappings into their granular counterparts in which the originally formed numeric prototypes are made granular so that the quality of the associative recall can be quantified. Several scenarios of allocation of information granularity aimed at the optimization of the characteristics of recalled results (information granules) quantified in terms of coverage and specificity criteria are proposed.

See more about Prof. Pedrycz here.

Dominik Ślęzak

Institute of Informatics, University of Warsaw, Poland

“Toward Approximate Analytics – Approximate Query Engines & Approximate Data Exploration”

Artificial Intelligence (AI) methods are regaining a lot of attention in the areas of data analytics and decision support. Given the increasing amount of information and computational resources available, it is now possible for intelligent algorithms to learn from the data and assist humans more efficiently. Still, there is a question about the goals of learning and a form of the resulting data-driven knowledge. It is evident that humans do not operate with precise information in decision-making and, thus, it might be unnecessary to provide them with complete outcomes of analytical processes. Consequently, the next question arises whether approximate results of computations or results derived from the approximate data could be delivered more efficiently than their standard counterparts. Such questions are analogous to the ones about precision of calculations conducted by machine learning and KDD methods, whereby various heuristic algorithms could be boosted by letting them rely on approximate computations. This leads us toward discussion of the importance of approximations in the areas of machine intelligence and business intelligence and, more broadly, the meaning of approximate derivations and representations for various aspects of AI. In this talk, we refer to this discussion using three industry-related case studies: 1) The case of approximate analytical database software based on the paradigms of rough-granular computing applied in the area of cyber-security; 2) The case of rough-set-based feature subset ensemble selection / approximation methodology applied in the area of online health support services; and 3) The case of approximate generation of the training data used for tuning an online eSports coaching platform.

See more about Prof. Ślęzak here.

Download Prof. Ślęzak short CV here.

Richard Weber

Professor, Universidad de Chile, Chile

“Dynamic Data Mining ”

We are witnessing a tremendous interest in data mining and related topics in research, industry, and public organizations. Virtually all areas of our daily life are affected, such as health, security, business, education, transportation, to name just a few. But most of the systems that are being used today are static in the sense that they consider snapshots of the respective phenomena under study. In this talk, we will present situations where dynamics play a crucial role in order to better understand the analyzed behavior. By reviewing some techniques for dynamic data mining we will provide the current state as well as challenges of this area. The talk ends with some ideas regarding future developments. 

See more about Prof. Weber here.

Pablo Zegers

“Artificial Intelligence, a Revolution in Latin America As Well”

The impact caused by Artificial Intelligence everywhere follows two main drivers: (i) the extent to which Artificial Intelligence (AI) can be applied in human society, a discipline that basically aims at building machines that mimic all human behavior, hence with the potential of affecting all human activity, (ii) the lucky discovery of mathematical rules that facilitate building incredible complex learning machines. These two drivers merged some 5 years ago and started what is now called the AI revolution. In this presentation these two driving components will be explained, followed by a review of the current state of the art.

Given that AI has been present in games, movies, and books, it is important to gain a precise knowledge of what can currently be achieved by this field, and to be conscious of its limitations, in order to avoid unrealistic expectations. 

Also, in order to understand how serious is what is happening, a review of the reactions of the private sectors, the government in many countries, and the societies all around the planet, is presented as well. The presentation continuous with an analysis of the expected scenarios, and how the job market is expected to adapt to the changes induced by the introduction of AI systems. The hybrid scenario, also called the centaur scenario, where man and machine work together, allowing for important quality and productivity increases, is analyzed.

Finally, the Latin American view is presented, where local opportunities are pointed out. In general, these opportunities are where: (i) processes are composed of very structured subprocesses, ripe for replacement by an AI system, (ii) private silos with private data exist, out of the reach of the big AI companies, where local companies can gain the trust of their local peers and offer AI services, and (iii) things need to run out of the internet, thus no SaaS system can compete with them (massive real-time video processing, AIs coordinating robots, etc.). The presentation continues pointing out the challenges in Latin America, mostly related to the lack of prepared professionals in the field, impeding a fast deployment of the technologies, deterring a fast product assembly, henceforth weakening the local capacity to compete with companies from abroad. The presentation ends with an analysis of the ethical implications of applying AI technologies in a society. 

Download Pablo Zegers CV here.