Société Luxembourgeoise de Statistique

15/01/2025

The Luxembourg Statistical Society is organizing its first LSS prize ceremony. This event will award for the first time a LSS prize for the best internship report for a master thesis of the University of Luxembourg.

The aim of this event is to reinforce the importance of statistics as an applied discipline and bring people together with a shared interest in applied statistics. It awards the students that did an exceptional work contributing to statistics via their internship report.

When and where? This event will take place on the 31st of January 2025 at 6 pm at Campus Belval (room to be announced soon)
Save the date!

Who are the nominees?

• Aleksandra Frania, Master in Data Science: Causal inference Methods and Tools (Internship done at Goodyear)

• Fatéma Goulamaly, Master in Industrial Mathematics: Predicting the Degradation of Li-ion Battery Using Hybrid Machine Learning and Empirical Methods (Internship done at Siemens Industry Software SAS)

• Rym Kaced, Master in Financial Mathematics: Insurance and ALM: Introduction to Strategic Asset Allocation (Internship done at PWC)

Program:

• 18:00: Welcome words by Christophe Ley, president of the LSS and introduction to the LSS prize by Yves Dominicy, scientific jury president.
• 18:10: Presentations by the Nominees
• 19:00: Jury Deliberation
• 19:15: Announcement of the Winner

If you are interested, please register to this event by sending an email to [email protected].

See you there,
LSS Board

04/12/2024

The registration for the MathSport International Conference 2025 taking place in Luxembourg from 4th to 6th June 2025 is now open.

Team MIDAS Luxembourg - Modeling Interdisciplinary Data (Science) Applied Statistics

04/07/2024

Thank you to all the members that joined us yesterday at the Summer Fest of the Luxembourg Statistical Society - Société Luxembourgeoise de Statistique at Domaine Mathis Bastian (Domaine Viticole).

Special thank you to Pierre Mangers from MANGHINI Consulting for the very interesting talk, and thank you to Annick Bastian for hosting us.

Great atmosphere, amazing location and delicious food!

The network of statisticians in keeps growing - what are you waiting to join us 😉

29/03/2024

We are happy to announce that the talk by Professor Masanobu Taniguchi from Waseda University, Japan will take place on Wednesday April 17 at 6.30pm at the University of Luxembourg on the Kirchberg Campus (6, Rue Richard Coudenhove-Kalergi) in the main building, room D10F. The talk will be followed by a drink.

Could you please inform us a soon as possible (if you have not done so already), if you would like to attend the event? (email: [email protected])

The title and abstract of the talk can be found here under.

Best regards,

LSS Board

--------------------------------------------------------

The title of the talk is:

Invitation to Time Series Analysis: Theory ＆Applications

And the abstract is the following:

"In this talk, topics from the elementary statistics to modern new one will be delivered. First, we explain the meaning of mean, variance, deviation, of statistical data. As the theoretical aspect, we introduce mathematical statistics, mentioning estimator, unbiasedness, sufficiency etc. For goodness of estimator, Rao-Blackwell theorem is stated with an interesting episode of C.R.Rao.
Then we explain the elements of time series, feature of financial data, and nonlinear time series models. Their optimal inference is discussed based on LAN structure, and as an optimal estimator, MLE is given. We apply time series analysis for medical data e.g., ECoG, EMG data to illuminate diseases.
Actual data are essentially non-stationary. For such time series, the optimal inference is established. Classification problems are addressed for such data.
We also deal with the problem of portfolio. For this the optimal portfolio coefficients are given by our modern statistical methods.
As an interesting episode, Nightingale’s life will be mentioned. We explain her relationship to statistics and graphical method in statistics.
Finally, Granger causality will be introduced two time series, and we apply this concept for tourism statistics."

19/03/2024

We are happy to announce our next event, which will be a talk by Professor Masanobu Taniguchi from Waseda University, Japan. The talk will take place on Wednesday April 17 (in the evening), details will follow, so save the date.

Could you please inform us a soon as possible, for organisational purposes, if you would like to attend the event? Send an email to [email protected].

The title of the talk is: Invitation to Time Series Analysis: Theory ＆Applications

And the abstract is the following:

"In this talk, topics from the elementary statistics to modern new one will be delivered. First, we explain the meaning of mean, variance, deviation, of statistical data. As the theoretical aspect, we introduce mathematical statistics, mentioning estimator, unbiasedness, sufficiency etc. For goodness of estimator, Rao-Blackwell theorem is stated with an interesting episode of C.R.Rao.

Then we explain the elements of time series, feature of financial data, and nonlinear time series models. Their optimal inference is discussed based on LAN structure, and as an optimal estimator, MLE is given. We apply time series analysis for medical data e.g., ECoG, EMG data to illuminate diseases.

Actual data are essentially non-stationary. For such time series, the optimal inference is established. Classification problems are addressed for such data.

We also deal with the problem of portfolio. For this the optimal portfolio coefficients are given by our modern statistical methods.

As an interesting episode, Nightingale’s life will be mentioned. We explain her relationship to statistics and graphical method in statistics.

Finally, Granger causality will be introduced two time series, and we apply this concept for tourism statistics."

27/02/2024

Dear LSS members,

Save the date: Webinar by Cristiano Tessitore from Eurostat which will take place Thursday 7th of March at 12:00pm.

To get the Webex link, please write an email to [email protected].

Here the title and abstract of the talk:

Title : “Large Language Models and Blockchain in Statistics: Innovations and Use Cases”

Abstract :
“This talk delves into the transformative and innovative potential of two cutting-edge technologies - Large Language Models (LLMs) and blockchain - for the field of statistics, and in particular, Official Statistics.

LLMs, with their enormous data processing and natural language understanding capabilities, have gained increasing popularity since the launch of OpenAI's ChatGPT in November 2022. This development has made the generative AI sector increasingly competitive, with new breakthroughs emerging daily, making LLMs a perfect “teammate” for scientists.

Blockchain technology, thanks to its innovative features like immutability, transparency, and decentralization, can strengthen trust in official statistical results.

By combining LLMs and blockchain, this speech aims at showcasing a future where statistical analysis is more efficient, accurate, and trustworthy. Attendees will gain insights into practical applications and the potential of these technologies to transform the statistical landscape.”

08/02/2024

Please find here a nice page (by LSS member Prof. Antonella Perucca) about some statistical traps with examples that can be interesting for the general public, students and statistics lovers

It is not that unusual, that something unusual ends up happening.

04/01/2024

Dear all,

Please find here the registration form for the LSS membership 2024.

Once filled in, it has to be returned to the email address [email protected] and the payment has to be done via the account given in the registration form.

Best regards,
LSS Board

12/09/2023

Tomorrow:

"Understanding the Potentials and Pitfalls of ChatGPT and AI Language Models" given by Raul Ian Sosa, University of Luxembourg.

When: September 13 at 6pm

Where: Metz Room, ground floor of the SnT Building (29, av, JF Kennedy, Luxembourg)

07/09/2023

Do not forget our next event is taking place next week September 13 at 6pm in the Metz Room, ground floor of the SnT Building (29, av, JF Kennedy, Luxembourg). It is a talk about "Understanding the Potentials and Pitfalls of ChatGPT and AI Language Models" given by Raul Ian Sosa, University of Luxembourg.

25/08/2023

LSS is pleased to announce its next event which is a talk by Raul Ian Sosa, University of Luxembourg.

The title of the talk is:

Understanding the Potentials and Pitfalls of ChatGPT and AI Language Models

Abstract: In recent months, AI language models have skyrocketed in popularity among both companies and end users. ChatGPT, in particular, has become the fastest-growing web platform of all time and serves as a formidable tool for tasks ranging from simple text summarization to complex ones such as writing entire emails and generating pieces of software. In this talk, we will delve into the creation and evolution of this revolutionary technology, demystifying what these AI models are and envisioning the future they might shape. Despite these new solutions offering immense possibilities, they also pose challenges such as the risk of generating misleading information and perpetuating societal biases. We will dissect these issues, understand their roots, and explore strategies that we, as users, can employ to mitigate them, ensuring the responsible use of these sophisticated tools.

Date and Place: September 13, 6pm, Metz Room, ground floor of the SnT Building (29, av, JF Kennedy, Luxembourg)

25/08/2023

Forwarding a message from LIH:

TRAINING: Statistical issues in drug development - Prof Stephen Senn - 12-13/09

We are pleased to announce a 2-day training with Prof Stephen Senn, Statistical consultant, taking place on 12th and 13th September at the Maison du Savoir (MSA) in Belval/Esch-sur-Alzette. He will give a series of interactive lectures on “Statistical issues in drug development” intended for non-statisticians who wish to gain an intuitive understanding of statistical issues related to clinical trial practice but also to statisticians who wish to understand practical problems that trialists face.

Please find more information in the attached picture.

LIH staff: free of charge – registration in MPLEO
Externals: 130 Euros (65 Euros 1 day) - registration by mail to [email protected]
Deadline: 4th September

19/07/2023

Looking for a community where you can connect with like-minded people and discuss all things analytics? Look no further than the Luxembourg Statistical Society (LSS)!!! Our community comprises economists, statisticians, data scientists, mathematicians, engineers, and more, all united by a passion for data. Whether you are a veteran or just starting out in the world of analytics, the LSS is the perfect place to network, share knowledge, and grow your skills.

13/04/2023

Reminder that our next event is happening Monday 17 April at 12pm online and is a talk by Prof. Monica Pratesi.

To get the Webex link, please write an email to [email protected].

30/03/2023

Save the date: Online seminar by Prof. Pratesi will be taking place Monday 17 April at 12:00pm

To get the Webex link, please write an email to [email protected].

Here the title and abstract of the talk:

Title : Towards a framework to harness data by citizens: the Istat experience

By Monica Pratesi, Director of the Department for Statistical Production, Istat, Italian National Institute of Statistics

Big data, smart statistics, digital administrative data, and citizens are inseparable: from smartphones, meters, fridges and cars to internet platforms, the data of most digital technologies is citizen data, that is the data of the citizens and on citizens.

Their importance and their involvement ask for a thorough rethinking of the relations between the citizens and the production of official statistics. In the session we will focus on examples of the reuse of data generated by citizens on themselves (in processes like using services, making telephone calls or when they populate administrative archives during their lives as students, workers, retired persons etc.) to produce Official Statistics.

Examples will use the statistical production process from the Permanent Population and Households Census, from the integration process of utility data files (e.g. electricity, gas) in the design of different activities, like for instance the Household Energy Consumption Survey or data collection on violence against women and on discrimination (LGBTQ+ population, Roma and homeless) conducted in cooperation with civil society organizations, also touching upon the issue of the reuse of privately-held-data in official statistics.

25/02/2023

Photo from our first event of the year which took place past Thursday, and was a seminar by Prof. Sarah Gelper who gave an interesting talk about "Estimating the Effect of Second Screening on TV Viewing using Mobile Diaries".

17/02/2023

Dear members,

The first LSS event will be a seminar by Prof. Sarah Gelper taking place Thursday 23 February at 12 :30pm. If you attend it in person it will be in the following room on the first floor of the Maison du Nombre, MNO-E01-0146030, on the University Campus Belval.

If you will attend it virtually, please contact us to get the link.

Please find underneath the title and abstract of the seminar.

Kind regards,
LSS board

Title : Estimating the Effect of Second Screening on TV Viewing using Mobile Diaries

By Sarah Gelper, Renana Peres, Mitch Lovett


People often use second devices, such as their mobile or tablet, while watching TV. Such "second screening" might enhance engagement and increase viewership but also risks distracting viewers and reducing interest in the show. Estimating the effect of second screening on repeat viewing is challenging for three reasons. (1) We need to observe behavior on two different devices (the data challenge); (2) the effect of second screening might vary with viewer and show characteristics (the heterogeneity challenge); and (3) second screening is highly endogenous behavior (the endogeneity challenge). This paper leverages large-scale individual-level data from mobile diaries of 1,702 US TV viewers on 2,755 primetime TV shows over 3-6 weeks in which viewers report both viewing and second screening. The data are augmented to account for a broad set ofmarketing influences such as show communications and promos. We characterize heterogeneity by four latent viewer classes (heavy viewers, sports enthusiasts, sitcom viewers, and light viewers) and use a causal forest analysis to estimate the heterogeneous effects of second screening on repeat viewing. Despite the possible risk of distraction, second screening does not harm actual TV viewing, and this effect is robust over various types of second screening activities and across viewer segments. Second screening impacts attitudes more than behaviors - unlike actual viewing, stated viewing intentions are affected right after the second screening experience. Specifically, viewing intentions increase when people use a second screen to communicate about the show, but decrease when they use the second screen for show-unrelated activities.