This is the story of my assorted attempts to do some good with machine learning. Through its telling, I’ll highlight several models of organizing social good efforts, describe half a dozen social good problems that would benefit from our community’s attention, and present both resources and challenges for those looking to do some good with ML.
Date: 4 June 2021 Time: 20:00-21:30 Place: Microsoft Teams Fee: Free Registration: please register by sending an email to Michela Bia (firstname.lastname@example.org)
It is a pleasure to invite you to our second LSS Presentation of 2021, to take place on May 18, from 8pm – 9:30 pm via Microsoft Teams.
The talk will be given by Prof. Stephen Senn. Stephen Senn is currently providing consulting for CCMS since he retired in April 2018 from his position of CCMS head. He joined LIH in 2011 and was previously a Professor in Statistics at the University of Glasgow (2003) and University College London (1995-2003). In addition to working as an academic he has also worked for the pharmaceutical industry in Switzerland and the National Health Service in England. He is the author of three books, Cross-over Trials in Clinical Research (1993 & 2002), Statistical Issues in Drug Development (1997, 2007) and Dicing with Death (2003). His expertise is in statistical methods for drug development and statistical inference. He consults extensively for the pharmaceutical industry.
The response to the COVID-19 crisis by various vaccine developers has been extraordinary, both in terms of speed of response and the delivered efficacy of the vaccines. It has also raised some fascinating issues of design, analysis and interpretation. I shall consider some of these issues, taking as my example, five vaccines: Pfizer/BioNTech, AstraZeneca/Oxford, Moderna, Novavax, and J&J Janssen but concentrating mainly on the first two. Among matters covered will be concurrent control, efficient design, issues of measurement raised by two-shot vaccines and implications for roll-out, and the surprising effectiveness of simple analyses. Differences between the five development programmes as they affect statistics will be covered but some essential similarities will also be discussed.
It is my pleasure to invite you to our first LSS Presentation of 2021, to take place on March 4, from 8pm – 9pm via Microsoft Teams . The talk will be given by Françoise Kemp from the University of Luxembourg. She got a Master in Applied Mathematics from the University of Trier in 2018, and is now doing a PhD in Systems Biomedicine at the Luxembourg Center for Systems Biomedicine, and is Member of the COVID-19 Taskforce Research Luxembourg
Against the current COVID-19 pandemic, non-pharmaceutical interventions have been widely applied; vaccines are also becoming available. Now, an urgent question is how the interplay between vaccination strategies and social measures will shape infections, hospital demand and casualties. Hence, we extend the Susceptible-Exposed-Infectious-Removed (SEIR) model to include vaccination, ICU, hospital and death. We calibrate it to data of Luxembourg, Austria and Sweden. The model quantifies the vaccination rate needed for herd immunity by desired times. Aiming to vaccinate the whole population within 1 year at a constant rate could lead to herd immunity only by mid-summer. Herd immunity might not be reached in 2021 if too slow vaccines roll-out speeds, which we quantify, are employed. Vaccination will help considerably, but not immediately, thus social measures still remain crucial for months.
National and Business Accounts for Sustainable Development Policies
The International Conference on Accounting & Assessment in Statistics (ICAAS) 2021 is a online conference, based on the overarching need to discuss ongoing developments on this topic in response to policy needs. ICAAS is held to address the issues of statistical methodologies and thematic applications in accounting and assessment. Discussions will focus on the new tools provision, particularly in the areas needed for sustainable development policies.
Several questions arise: which kind of information is needed for Sustainable Development? To what extent these information needs can be met by accounting in statistics? Which data are already available and which missing data should be collected? Which are the statistical issues to be addressed and the limits of this approach? Which are the best experiences and practices developed in this area? These and further issues in methodological and applied statistics: contributions are expected on both these areas.
The Luxembourg Statistical Society (LSS) is organizing the event with the support of FENStatS, Statec, Eurostat, University of Luxembourg, PWC-Luxembourg and the IARNIW of India.
Date: Monday 31 May to Friday 4 June 2021 Place: online Fee: free
It is our pleasure to announce that the International Conference on Statistics and Related Fields (ICON STARF) will take place between 12 and 16 July 2021 at the University of Luxembourg, on Belval campus. The conference will cover the fields of mathematical statistics, signal processing, statistical learning, probability, approximation theory, and other areas that contribute to the mathematical development of data science.
Our aim is to bring together young researchers and experts of these fields. Lectures will be given by leading international researchers from all over the world (China, France, Germany, the Netherlands, the United Kingdom, the United States, etc.). We shall allow time for discussion and stimulate scientific exchanges among participants. Registration fees are low and financial support for travel and accommodation is available for doctoral students, post-docs and young researchers.
ICON STARF international conference is an initiative of the SanDAL project funded by the European Union’s Horizon 2020 research and innovation programme under grant agreement No 811017. The University of Luxembourg has been awarded the SanDAL project to set up a new ERA Chair in mathematics and statistics.
Date: 12-16 July 2021 Time: 09:00-16:00 Place: University of Luxembourg, Belval Campus, Maison du Savoir 2, avenue de l’Université, L-4365 Esch-sur-Alzette Fee: 60€ Registration: please register before 1 June 2021 via this link
On Thursday, 10 December 2020, at 13:00, the Luxembourg Statistical Society is proud to welcome Michela Bia for a seminar on « The long-run effect of childhood poverty and the mediating role of education ».
Ms Michela Bia is secretary of the LSS and she works as researcher at LISER.
The manuscript, co-authored with Luna Bellani, has been published last year on the Journal of the Royal Statistical Society (Statistics in Society) Series A, 2019, vol. 182, n°1, pp. 37-69 .
The seminar will be held online and the link to the presentation wills provided on request.
The study examines the role of education as a causal channel through which growing up poor affects the economic outcomes in adulthood in the European Union. We apply a potential outcomes approach to quantify those effects and we provide a sensitivity analysis on possible unobserved confounders, such as child ability. Our estimates indicate that being poor in childhood significantly decreases the level of income in adulthood and increases the average probability of being poor. Moreover, our results reveal a significant role of education in this intergenerational transmission. These results are particularly relevant for Mediterranean and central and eastern European countries.
In the frame of the « ERA Chair in Mathematical Statistics and Data Science – SanDAL », the University of Luxembourg organises an online Winter School on Mathematical Statistics from 7 to 11 December 2020.
This school is aimed at young researchers (PhD students and early postdocs) and practitioners alike and offers an excellent opportunity to learn about recent and important developments in mathematical statistics directly from leaders in their respective fields
The following three courses will be offered:
Non-parametric Inference under Local Differential Privacy Cristina Butucea (CREST ENSAE, Université ParisTech) Principal Component Analysis: some recent results and applications Karim Lounici (CMAP-Ecole Polytechnique) Statistical inference of incomplete data models to analyse ecological networks Stéphane Robin (AgroParisTech/INRA/univ. Paris Saclay & Muséum National d’Histoire Naturelle)
Pierre Mangers, membre SLS, vous propose le graphique ci-dessus décrivant le temps de doublement du nombre de personnes testées positives à partir d’un effectif de 100 personnes. Le graphique s’appuie sur les données officielles publiées pour une partie de la Grande Région. Plus la courbe est plate, plus le temps de doublement des personnes testées positives est court, donc plus la propagation du virus est élevée. A l’inverse pour le Luxembourg, la courbe descend plus vite traduisant une propagation plus lente du virus. Il montre aussi qu’il y a une différence de 7 jours à la date de mi-avril entre le Luxembourg et la France.
Les dernières statistiques disponibles sur l’épidémie Covid-19 en date du 4 Mai. Le Ministère de la Santé a publié des nouvelles données sur les site open data. Les graphiques 1 et 2 comparent les deux séries, les autres graphiques sont réalisés avec l’ancienne série.
La courbe du nombre de personnes testées positives prend maintenant la forme d’une courbe logistique. On s’éloigne donc d’une croissance exponentielle.
S’il est vrai que le nombre absolu de personnes testées positives est très important, notamment pour les hôpitaux, le taux de variation journalier (total cas positifs par rapport au total du jour précédent) fournit une meilleure indication sur la propagation du virus. On voit bien que la croissance ralentit depuis la mi-mars et avec la régression du virus elle va tendre vers 0.
Ce graphique permets de comparer les taux de croissance des cas testés positifs de certains pays européens.
Le graphique 4 montre le nombre de nouveaux cas testés positifs journaliers. La ligne noire est une régression locale et représente la tendance.