UNIGE Data Science Day
The UNIGE Data Science Day is a scientific symposium for researchers at the UNIGE, which takes place at the beginning of the academic year.
Each year, the CCSD stimulates a collective reflection on a particular theme. This reflection is introduced during the UNIGE Data Science Day and continued in the framework of other activities launched by the Center throughout the year. This theme is intended to be precise enough to encourage a significant scientific contribution and a rich dialogue, but also transversal enough to allow for interdisciplinarity.
For its 2022 edition, the theme of Data Science Day is "Promises of Artificial Intelligence: An Interdisciplinary Revolution".
This meeting will take place on September 15, 2022 from 8:45 am to 6:30 pm.
More information on the programme below.
2022 Theme: Promises of AI
Artificial intelligence (AI) is synonymous with the greatest promises and the greatest challenges for humanity. A constant object of imagination and fantasy, AI featured in many popular works long before it became a reality, works projecting dystopian and utopian futures of what could become a world governed by a superior intelligence that would exceed the capabilities of its creator. Today, AI has left the realm of mere intellectual projection to become an integral part of our digital landscape. There are traces of it everywhere: in the algorithm of an online shopping website, in facial recognition applications, or even behind a comment on a social network. Continuous rise in computing power combined with a considerable increase in the amount of computer data now available is what has made this transformation possible, by offering the capacity to train computers to learn and thus to develop an intelligence of their own. In the sciences, these new capabilities offer so many possibilities for rapid exploration and data processing that they can be seen as a truly promising revolution.
For this third edition of the UNIGE Data Science Day, the Data Science Competence Centre of the University of Geneva (CCSD) invites researchers from the University of Geneva to take part in an interdisciplinary study day on artificial intelligence and its various scientific applications. This meeting will be held on 15 September 2022 at Uni Mail. In order to embrace the diversity of the work carried out at the UNIGE on this theme, this day will be articulated around three dimensions of AI:
- AI and the fundamental knowledge on which its development is based,
- AI as an applied tool in the context of a growing number of research projects, and
- AI as a research object in itself, which must be approached in a critical manner.
Within the first dimension, the CCSD would like to invite UNIGE researchers working on the development of AI at a fundamental level to present their work. Whether it focuses on the creation of artificial intelligence architectures, the improvement of machine learning algorithms or the improvement of the technical tools used, all research on this topic (in particular in computer sciences and statistical sciences) is welcome. For the second dimension, the CCSD invites UNIGE researchers to present the multiple applications of artificial intelligence in the framework of their research, regarding the use of image, text, sound or graph processing techniques for classification, categorization or prediction. Finally, the third dimension is open to researchers working to better understand the social, political, economic, legal and ethical implications of the advent of AI and its growing use in society and in scientific circles.
Following the study day on 15 September 2022, the CCSD and the Scientific Computing Support (SciCoS) will launch a joint call for projects for an impulse fund awarding a one-year grant of CHF 20,000 to the best interdisciplinary project on the topic of AI.
8h45-9h00: Introduction (room MR280)
- Guive Khan Mohammad, Digital Transformation Office, "UNIGE Data Science Competence Center"
- Elsa-Line Huwyler, Data Science Competence Center, "Data Science for All"
9h00-11h00: Plenary Session "Promises of AI: An Interdisciplinary Revolution" (room MR280)
- Tobias Golling, Faculty of Sciences, Particle Physics Department, "The fifth paradigm of scientific discovery"
- François Fleuret, Faculty of Sciences, Department of Computer Science, "Understanding through generative modeling"
- Svyatoslav Voloshynosky, Faculty of Sciences, Department of Computer Science, "Challenging traditional astronomy imaging pipelines"
- Giovanna Di Marzo Serugendo, Geneva School of Social Sciences, Computer Science Center, "Multi-agents systems - a tool for modelling and developing complex systems"
- Jose Luis Fernandez-Marquez, Computer Science Center, "TBD"
- Stefan Sperlich, Geneva School of Economics and Management, Institute of Economics and Econometrics, "Semi-soft methods : Machine Learning you can understand and infer on"
- Paola Merlo, Faculty of Humanities, Linguistics Department, "Disentangling linguistic intelligence: learning structure and meaning across languages"
- Nicola Carboni, Faculty of Humanities, Digital Humanities, "Images in Global Circulation: A Multi-scalar Approach"
- Giuseppe Ugazio, Geneva School of Economics and Management, Geneva Finance Research Institute, "The role and needs of philanthropy to promote EIAI”
- Katarzyna Wac, Geneva School of Economics and Management, Information Science Institute, "AI for Quality of Life Assessment: Where is the Human?"
- Marc Audard, Faculty of Sciences, Department of Astronomy, "AI at the Department of Astronomy"
- Luca Caricchi, Faculty of Science, Department of Earth Sciences, "Unsupervised and supervised machine learning applied to volcanology"
- Denisa Rodila, Institute for Environmental Studies, Spatial Predictions and Analyses in Complex Environments, "AI applications on Environmental Sciences"
- Xin Wen, Institute for Environmental Sciences, Renewable Energy Systems group, "Using retrospective modeling to inform choices in developing bottom-up electricity system models"
- Christian Lovis, Faculty of Medicine, Medical and Information Sciences, "Medicine : a multidisciplinary science requiring a multidisciplinary approach"
- Patrycja Nowak-Sliwinska, Faculty of Medicine, Translational Research Centre in Oncohaematology, "AI for optimization of multidrug combinations"
- Dino Vajzovic, Faculty of Law, Department of Civil Law, "Swiss liability law in the face of digital manipulation"
- Isabelle Collet, Faculty of Psychology and Educational Sciences, Gender and Intersectional Relations Research Group in Education, "Hey Siri, why do you speak with a woman's voice?"
- Steven Schramm, Faculty of Sciences, Particle Physics Department, "Artificial Intelligence: Bringing Together Industry, Science, and Society"
11h00-12h30: posters session AI (Hall of Uni Mail)
- Flann Chambers and Giovanna Di Marzo Serugendo, Computer Science Centre, "TRACES - l’IA pour l’analyse de la trajectoire des territoires"
- Philippe Glass and Giovanna Di Marzo Serugendo, Computer Science Centre, "Lasagne - implementation of smartgrid in the form of a coordination system of digital twins"
- Atul Sinha, Computer Science Centre, "Fast Detector Simulation and Detector Design"
- Vitaliy Kinakh, Computer Science Centre, "MV-MR: multi-views and multi-representations for self-supervised learning based on dependence maximization"
- Balint Mate, Computer Science Centre, "Flowification"
- Amudha Ravi Shankar, Computer Science Centre, "Collective Intelligence for Crisis Response"
- Hafiz Budi Firmansyah, Jose Luis Fernandez-Marquez and Jesús Cerquides, Computer Science Centre, "Artificial Intelligence Approach To Improve Social Media Images Classification For Crisis Response"
- Akouété Antonin Sedoh and Ambroise Barras, service Culture de l’UNIGE, "Cultura : agent conversationnel du service Culture de l’UNIGE"
- Adrien Jeanrenaud, Faculty of Humanities, "How to recognize a "film noir"? A digital approach to film genres through posters (1945-1990)"
- Dino Vajzovic, Faculty of Law, "Can swiss liability law protect you against digital manipulation ?"
- Julien Prados, Faculty of Medicine, "Getting AI support at bioinformatics support platform for data analysis"
- Benoit Girard, Faculty of Medicine, "Computational approaches for social ethology in rodents"
- Jean-Philippe Goldman, Faculty of Medicine, "Hybrid approach for de-identification of texts"
- Mina Bjelogrlic, Faculty of Medicine, "HERO - Human Extraordinary Robust Organism : data-driven research on protective factors"
- Hugues Turbé, Faculty of Medicine, "Interpretable models for time series classification"
- Jamil Zaghir, Faculty of Medicine, "Performance of Machine Learning Methods to Classify French Medical Publications"
- Daniel Keszthelyi, Faculty of Medicine, "Summarizing patient report for healthcare professionals: combining data-driven and rule-based approaches"
- Belinda Lokaj, Faculty of Medicine, "Do better with less : Combining ultrafast MRI with artificial intelligence for breast cancer detection"
- Alexandra Villaverde Naveira, Faculty of Medicine, "Robotics devices as digital solutions to improve medical care on elderly"
- Christophe Gaudet-Blavignac, Faculty of Medicine, "CoviDB, a SNOMED CT enabled database for COVID-19 related data"
- Christophe Gaudet-Blavignac, Julien Ehrsam and Deniz Geres, Faculty of Medicine, "Health data semantics in the Swiss Personalized Health Network: a Three-Pillar Strategy"
- Berry Holl, Lorenzo Rimoldini, Panagiotis Gavras, Krzysztof Nienartowicz, Laurent Eyer, Nami Mowlavi, Grégory Jeverdat De Fombelle and Marc Audard, Faculty of Sciences, "Toward big data classification of variable sources from a trillion observations by the Gaia spacecraft"
- Corin Jorgenson, Faculty of Sciences, "Using machine learning techniques to determine pressure and temperature of magma storage below volcanoes"
- Samuel Klein, Faculty of Sciences, "Pulling back the CURTAINs on new Physics (CURTAINs)"
- Guillaume Quetant, Faculty of Sciences, "Transcoding between two distributions with GANDALF"
- Debajyoti Sengupta, Faculty of Sciences, "Fantastic new physics and how to find them (Anomaly Taggers with CLR)"
- Matthew Leigh, Faculty of Sciences, "nu-Flows: conditional neutrino regression"
- Malte Algren, Faculty of Sciences, "Flavour tagging calibration using optimal transport"
- Olga Taran, Omkar Bait, Miroslava Dessauges, Svyatoslav Voloshynovskyy and Daniel Schaerer, Faculty of Sciences, "Challenging radio-astronomy imaging: a direct source localization from uv-plane observations"
- Vincent Micheli, Faculty of Sciences, "Transformers are sample-efficient world models"
- Danièle Paliotta, Faculty of Sciences, "Transformers and Graph Neural Networks"
- Julia Roquette, Faculty of Sciences, "The Orion star-formation complex as a training set for machine learning techniques"
- Michael Fuchs, Faculty of Sciences, "Recognition of social gestures and body signals among great apes using video analysis"
- Krzysztof Nienartowicz, Sednai, ScienceNow, Laurent Eyer, Faculty of Sciences, "Gaia Vari: Designing a machine learning worflow with Citizen Science in the loop"
- Vitaliy Kinakh, Taras Holotyak and Svyatoslav Voloshynovskiy, Faculty of Sciences, "Hubble meets Webb: AI based image-to-image translation"
- Pierrette Bouillon, Bastien David, Johanna Gerlach, Lucia Morado Vazquez, Johny Mutal, Lucia Ormaechea Grijalba, Silvia Rodriguez Vazquez, Marianne Starlander, Irene Strasly, Nikos Tsourakis, Faculty of Translation and Interpreting, "Multilingual applications for positive social impact"
- Igor Almeida Matias, Geneva School of Economics and Management, "Providemus alz: Giving you the power to foresee Alzheimer’s disease effortlessly"
- Clauirton De Albuquerque Siebra, Geneva School of Economics and Management, "Explainable AI for Longitudinal Behaviour, Health and Quality of Life Data"
- Katarzyna Wac, Geneva School of Economics and Management, "Daily Life Sensing for Quality of Life Assessment: mQoL Living Lab"
- Rahul Kumar Jha, Geneva School of Economics and Management, "RIP Customer Satisfaction! Measure quality as experienced by Customers for better diagnostic of your product"
- Gregory Giuliani, Institute for Environmental Studies, "Using AI to map Switzerland's Land Cover using time-series of satellite data"
- Anthony Lehmann, Institute for Environmental Studies, "enviroSPACE: Spatial Predictions and Analyses in Complex Environments"
- Stéphane Paltani, Alejandro Alvarez Ayllon, Guillaume DEsprez, Florian Dubath, William Hartley, Nicolas Morisset, Federica Tarsitano and Marco Tucci, Faculty of Sciences, "The Euclid Photomeric Redshift Pipeline"
- Simon Gabay, Ariane Pinche and Kelly Christensen, Faculty of Humanities, "Gallic(orpor)a: Processing Gallica's Historical Sources"
- Division de l'information scientifique, Research Data Team
- Division Système et Technologies de l’Information et de la Communication, e-research Team
- Presentation of SmartLab project from the Medical Faculty : Laboratory Information Management System (LIMS) implementation, Coralie Fournier
12h00-13h30: lunch (Hall of Uni Mail)
13h30-15h00: first panels / rounds tables session
Organised by Béatrice Joyeux-Prunel, Tommaso Venturini and Simon Gabay
Nicola Carboni (University of Geneva, Faculty of Humanities), Jean-Luc Falcone (University of Geneva, Computer Science Centre), Simon Gabay (University of Geneva, Faculty of Humanities), Béatrice Joyeux-Prunel (University of Geneva, Faculty of Humanities) and Tommaso Venturini (University of Geneva, Geneva School of Social Sciences).
For too long, the humanities and social sciences have been bogged down in binary oppositions such as close VS distant reading or qualitative VS quantitative research, which, despite their historical raison d'être, have ended up anesthetizing our methodological imagination. Time has come to be more creative and to seize the renewal opportunities offered by the rise of digital records and computational tools to overcome the old quali/quanti opposition and try out newer and more creative research methods.
This event is dedicated to the methodological creativity spurred by digital methods. It offers a safe space to discuss playfully, but also very seriously, about new and different research protocols to study human and social phenomena. If you have a research method that feels unorthodox, a quirky dataset, and offbeat technique of analysis or visualization, this event is for you.
Organised by Stefan Sperlich
Olivier Renaud (University of Geneva, Faculty of Psychology and Education Sciences), Frank Röttger (University of Geneva, Geneva School of Economics and Management) and Stefan Sperlich (University of Geneva, Geneva School of Economics and Management)
Data are often used to study causal relationships, and consequently, empirical causal analysis has been an important topic in many different domains among which medicine and biology, including psychology were probably the first. While in some disciplines, experiments are the obvious way to go, causal analysis is more involved where nonexperimental data are to be used. In any case, the choice of one of the available methods should be problem- and data driven. In some cases, also the application of different methods is thinkable, as long as one bears in mind that they answer slightly or sometimes quite different questions (i.e., solve different problems). This session discusses in three separate presentations different aspects of smart data-driven causal analysis, where ‘smart’ refers mainly to a statistical modelling driven by domain knowledge. We also introduce graph-theory as a useful approach for structuring and choosing models and methods. It is particularly attractive as on one side it has a sound mathematical base but on the other hand can be practically applied in essentially any discipline. Other aspects discussed are for example unexpected complications in experiments as well as interpretation issues and limitations of causal analysis.
Organised by Tobias Golling and Steven Schramm
Samuel Klein (University of Geneva, Faculty of Sciences), Vincent Micheli (University of Geneva, Faculty of Sciences), Guillaume Quétant (University of Geneva, Faculty of Sciences) and Naoya Takeishi (University of Geneva, Faculty of Sciences)
Automated prediction is one of the major drivers of modern machine learning, be it weather prediction, protein folding or the prediction of readout signals in a measurement device. Generative modelling techniques have become the gold standard for a wide range of applications, with the most popular models including variational auto encoders (VAEs), generative adversarial networks (GANs), normalising flows or diffusion models. The automated detection of anomalies builds on this precise modelling of high-dimensional data distributions through generative modelling. Applications range from the search for the largest objects in space to the tiniest particles in the collisions of the Large Hadron Collider at CERN. In this panel, we intend to focus on early career scientists, inviting them to talk about their challenges and the proposed solutions, as well as how the approaches could be generalised.
Organised by Giuseppe Ugazio
Huber Halope (WEF-GAIA), Rahul Kumar Jha (UN-ITU), Milos Maricic (The Altruistic League) and Giuseppe Ugazio (University of Geneva, Geneva School of Economics and Management), Katarzyna Wac (University of Geneva, Geneva School of Economics and Management)
This panel/round table discusses the ways in which AI has been / can be used for pro-social purposes. AI has demonstrated its unique potential for aiding businesses, research institutions, and other organizations. A growing movement including Non-profit and International Organisations based in Geneva are pushing for the deployment of AI for both a) to promote social good, and in particular aiding several initiatives aimed at accelerating the achievement of the SDGs, and b) to leverage the potential of AI technology for its own increased operational efficiency. The goal of this panel is to discuss where we stand with respect to the adoption of AI for promoting social good, with a particular focus on discussing what are the most promising ways in which AI can be used by organizations working to promote public good, such as philanthropic and/or international organizations such as ITU.
15h30-17h00: seconde panels / rounds tables session
Organised by Paola Merlo and Emilie Wyss
Valentina Borghesani (University of Geneva, Faculty of Psychology and Education Sciences), James Henderson (IDIAP), Itsaso Olasagasti (University of Geneva, Faculty of Medicine) Nikhil Phaniraj (University of Zurich, Department of Anthropology) and Martin Volk (University of Zurich, Department of Computational Linguistics)
Artificial intelligence has a significant impact on the evolution of human language and our communication. Machines that create (fake) news, conversations with our voice assistant or brain-computer interfaces that predict our speech are no longer pipe dreams but a reality. What does this mean for the evolution of language? Where are we heading to with this? Where do we want to arrive?
This round table is composed of specialists from diverse but essential disciplines for the study of language. They are all part of the National Centre of Competence in Research “Evolving Language”, which has set itself the goal of deciphering the origins of language, its evolution over time and its future. The discussion will show that through the lens of the study of language, artificial intelligence is not only an object of research and a catalyst for change but also a tool for innovative and promising research.
Organised by Marta Pittavino
Pierre Bonnaventure (University of Geneva, Faculty of Medicine), Denis Chastagnier (University of Geneva, Faculty of Medicine), Jiahao Chen (University of Geneva, Geneva School of Economics and Management), Elsa Giobellina (University of Geneva, Animal Experimentation), Fanny Lebreton (University of Geneva, Faculty of Medicine), Jean-Luc Pitetti (University of Geneva, Faculty of Medicine) and Daniele Roppolo (University of Geneva, Animal Experimentation)
When dealing with animal experimentation, one of the first aims is to work with the right number of animals, which for ethical reasons should be the smallest possible while keeping the statistical significance and power:
- Q1) How can the data science help us REDUCE the number of animals? How to calculate the number of animals needed for an experiment when not all information is available, for example in case of exploratory study? How to reduce not only the number of animals used in experiments, but also the number of animals crossed and born in animal facilities?
Pre-clinical studies with animals are sometime seen as the cause of the failure of clinical trials and part of the general “reproducibility crisis” in research. Exclusion of a sex in experiments with animals may give false hint when translating results to men and women; however, exclusion may be justified in some experimental models.
- Q2) How can the data science lead to REPRODUCIBLE RESULTS of existing studies with a more balanced layout? How data science can help researchers including as often as possible both sexes in order to obtain reproducible and translatable results, while keeping low the number of animals used low?
Last, but not least, a primary instrument when conducting research with animals is the monitoring of animal welfare by completion of dedicated clinical score sheets. A clinical score sheet or “welfare assessment protocol” was established by Morton and Griffiths in 1985 (Morton and Griffiths, 1985) as a tool to grade the suffering of animals and to determine humane endpoints. Retrospective analysis of score sheets is not commonly used and practiced in the animal experimentation domain.
- Q3) How can data science be used to REFINE further animal experimentation by using information from the “score sheet”?
The three questions above will be addressed and discussed by scientists and experimenters during the above-mentioned round table.
Organised by Lamia Friha and Giovanna Di Marzo Serugendo
Alexander Barclay (OCSIN), Cristina Bueti (UIT), Philippe Glénat (Firmenich), Frédéric Joss (SIG) Reymond Laurent (OCSIN), Christian Pellegrini (University of Geneva), Christophe Salères (Rolex) and Laura Tocmacov Venchiarutti (ImpactIA).
Moderator: Vasu Briquez
Swiss companies and in particular SMEs must adopt a digital transformation in order to remain competitive, resilient and autonomous in relation to technologies. The degree and measurement of the level of digital maturity is variable and highly dependent on the domain and may or may not require the adoption of advanced technologies or even artificial intelligence. According to the SME portal of the Swiss confederation, digitalisation consists of changing the business model by taking into account advanced IT. The turn towards technologies potentially using AI can range from the adoption of an intelligent service (chatbot) for customer relations, to the exploitation, analysis and valorisation of knowledge using large quantity of data available to SMEs (Big Data, Data Science, etc.) or improving the efficiency of business processes. This round table brings together several experts involved in digital transformation and strategy within public, private, international organisations in the canton of Geneva.