The fields of machine learning and pattern recognition can arguably be considered as a modern-day incarnation of an endeavor which has challenged mankind since antiquity. In fact, fundamental questions pertaining to categorization, abstraction, generalization, induction, etc., have been on the agenda of mainstream philosophy, under different names and guises, since its inception. As it often happens with scientific research, in the early days of machine learning there used to be a genuine interest around philosophical and conceptual issues, but over time the interest shifted almost entirely to technical and algorithmic aspects, and became driven mainly by practical applications. With this reality in mind, it is instructive to remark that although the dismissal of philosophical inquiry at times of intense incremental scientific progress is understandable to allow time for the immediate needs of problem-solving, it is also sometimes responsible for preventing or delaying the emergence of true scientific progress.
With the increasing pervasiveness of digital data in all aspects of human life and the advent of the so-called “data revolution,” moreover, a new lively discussion is taking place concerning the societal and ethical implications of machine learning and other big-data enabling technologies. A considerable literature has accumulated on the nature and the value of data-driven technologies in human life and for the sake of society. This involves a variety of ethical and legal issues traditionally associated to the development of ICT and is now exacerbated by such controversial topics as, e.g., privacy, confidentiality, ownership and online identity. But the ethical discussion includes also the effects of data analytics on the development of culture and research (e.g., Will large quantities of data positively transform how we study and gain knowledge? Is data-driven science promoting new forms of radical empiricism?), in particular with respect to social science and the humanities (Will data analytics improve our understanding of social interactions? Are big data re-inscribing established divisions between scientific method and humanistic inquiry?), as well as the creation of new services and business circles.
All this suggests that the time is ripe to attempt establishing a long-term dialogue between the philosophy and the pattern recognition and machine learning communities with a view to foster cross-fertilization of ideas. The goal of this tutorial is to provide a timely and coherent picture of the state of the art in the field and to stimulate a discussion and a debate within our community. We do feel that this could be an opportunity for reflection, reassessment and eventually some synthesis, with the aim of providing the field a self-portrait of where it currently stands and where it is going as a whole, and hopefully suggesting new directions. Interestingly the epistemological and the ethical issues can be viewed as two sides of the same coin as the development of learning algorithms requires, as pointed out by Wiener as early as the mid-fifties and in-depth analysis of the design principles acceptable to us. On the other hand, reflecting on the epistemological underpinnings of our discipline can in fact reinforce the idea that there is no isolated “pure” science and that scientific and technological progress is in most of the cases socially and ethically mediated.
We shall assume no pre-existing knowledge of philosophy by the audience, thereby making the tutorial self-contained and understandable by a non-expert.
Marcello Pelillo is Full Professor of Computer Science at Ca’ Foscari University of Venice, where he directs the European Centre for Living Technology (ECLT). He held visiting research positions at Yale University, McGill University, the University of Vienna, York University (UK), the University College London, and the National ICT Australia (NICTA). He has published more than 200 technical papers in refereed journals, handbooks, and conference proceedings in the areas of machine learning, pattern recognition and computer vision. He serves as a General Chair for ICCV 2017 and has initiated several conference series, including EMMCVPR in 1997, IWCV in 2008, SIMBAD in 2011. He serves (has served) on the Editorial Boards of several journals including IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), IET Computer Vision, Pattern Recognition, Brain Informatics, and he serves on the Advisory Board of the International Journal of Machine Learning and Cybernetics. He has served (serves) as Guest Editor for various special issues of IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), Pattern Recognition, IEEE Transactions on Neural Networks and Learning Systems (forthcoming). His interest in the philosophical aspects of machine learning and pattern recognition has recently led him to run several initiatives devoted to this topic, among which a workshop at NIPS in 2011, a special issue of Pattern Recognition Letters (2015), and a tutorial at ICPR 2014. He is (or has been) scientific coordinator of several research project such as SIMBAD, an EU-FP7 project devoted to similarity-based pattern analysis and recognition whose activity is described in a recently published Springer book, and he has recently won an award from the Samsung Global Research Outreach (GRO) program. Prof. Pelillo is a Fellow of the IEEE and a Fellow of the IAPR. He has recently been appointed IEEE SMC Distinguished Lecturer.
Teresa Scantamburlo is a post-doctoral researcher at the Department of Environmental Sciences, Informatics and Statistics (DAIS) at the university of Ca' Foscari (Venice) where she got a bachelor degree in computer science and a master degree in computer science for the humanities. In 2014 she received a PhD in Computer Science under the supervision of professor Marcello Pelillo and since 2015 she has been working in collaboration with the European Centre for Living Technology (ECLT) in Venice. Her research interests lie at the intersection of philosophy and pattern recognition/machine learning. Currently she is dealing with the socio-ethical issues of data science and its disparate application on human decision-making. Other research topics include the foundation of pattern recognition from the standpoint of the philosophy of science and the theory of knowledge, and the computational models of categorization in cognitive psychology.