Scientific Data Services

Talk on "Breaking up is hard (to model): Joint models for longitudinal and time-to-event data on divorce data and rare events"

2025-11-13T14:07:38+01:00

"Breaking up is hard (to model): Joint models for longitudinal and time-to-event data on divorce data and rare events" is the title of Sophie Potts talk, which is being organised by the "Zentrum of Statistik" at Bielefeld University. It will take place on Tuesday, 18 November 2025 from 12:00 - 13:00 in W9-109.

In time-to-event analyses in social sciences, endogenous time-varying variables often occur, where the event status is correlated with the covariate’s trajectory. Ignoring this causes biased estimates. In biostatistics, this is addressed with joint models for longitudinal and time-to-event data, which handle endogeneity correctly. Despite their usefulness, joint models remain rare in social sciences. We give an introduction to this method, highlight its advantages for social sciences, and demonstrate it with an example on marital satisfaction and dissolution, comparing results with classical time-to-event models. In a second part, we address rare events in joint models. Standard estimation struggles with few events and imbalanced designs, leading to monotone likelihood. Firth’s correction, which adjusts the score function, provides stable estimates even with very small event counts. We extend this correction to joint models, enabling their use in rare event studies, particularly in observational data where event numbers cannot be controlled.

Talk on "The more, the merrier: the role of exogeneity in global state space models"

2025-10-28T14:16:53+01:00

"The more, the merrier: the role of exogeneity in global state space models" is the title of Prof. Dr. Dietmar Bauer talk, which is being organised by the "Zentrum of Statistik" at Bielefeld University. It will take place on Tuesday, 4 November 2025 from 12:00 - 13:00 in W9-109.

Often multivariate time series data has matrix time series structure such that at each point in time the same variables are observed for a number of individual entities. Joint unrestricted modelling using vector autoregressions (VAR) contains a prohibitively large number of parameters. This can be countered in a number of fashions (for example, MaTS, aDFM, Bayesian VARs). These approaches, however, do not include structural information on the relations between the various individuals. The global VAR approach of Pesaran and co-workers provides a different idea: It partitions the model into regional models, each of which is estimated and specified separately, allowing for the imposition of structural relations. Partitioning uses heavily different exogeneity concepts. In this talk we examine the implications of these exogeneity concepts. We also propose how to proceed, if exogeneity is violated.

Talk on "Data-Driven Approaches for Reducing Food Waste in Grocery Retailing"

2025-10-14T14:39:28+02:00

"Data-Driven Approaches for Reducing Food Waste in Grocery Retailing" is the title of Dr. David Winkelmann talk, which is being organised by the "Zentrum of Statistik" at Bielefeld University. It will take place on Tuesday, 21 October 2025 from 12:00 - 13:00 in W9-109.

Grocery retailers face the trade-off between overstocking, potentially generating food waste, and risking customer dissatisfaction through stock outs. To ensure high service levels, many retailers accept excess inventory, contributing to the approximately 11 million tons of food discarded annually in Germany, over 10% of which originates in the retail sector. In collaboration with the German government, major retailers have therefore committed to halving food waste by 2030. Achieving this goal requires more accurate demand forecasts that capture operational complexities. While demand forecasting has been widely studied by the scientific literature, specific challenges remain. For example, retailers frequently apply price reductions to units nearing expiration: such discounts stimulate short-term customer demand and mitigate the risk for food waste, but if these inflated sales figures are not modelled appropriately, forecasts systematically overestimate future demand and perpetuate overstocking. Using data from a major European grocery retailer, this study aims to contribute to the reduction of food waste in the retail sector by improving forecasts for these specific challenges.

“Week of Research Competence (WoFoKo)” at University Bielefeld

2025-10-04T00:02:47+02:00

We are pleased to announce that the Week of Research Competence (Woche der Forschungskompetenzen WoFoKo) will take place again this year, for the third time.

Organized by the Kompetenzzentrum Forschungsdaten, the Servicezentrum Medical Data Science and the Scientific Data Services (SDS), the WoFoKo focuses on strengthening essential skills for (empirical) research for early-career researchers from all disciplines at Bielefeld University.

The WoFoKo will take place online via Zoom from the 6th to 10th of October 2025, between 09:15-15:00.

Key Topics are:

- Project Management & Conflict Management

- Data Security & Information Security

- Rights of Data

- Ethics of Data Collection

- Research Data Management

- Academic Writing & Publishing

- Open Educational Resources

- Data Literacy & Open Data

- Research Funding

- Healthy Research Practices

- Artificial Intelligence in Research

Talk on "New directions for knowledge tracing models"

2025-07-03T20:46:13+02:00

"New directions for knowledge tracing models" is the title of Prof. Dr. Benjamin Paaßen talk, which is being organised by the "Zentrum of Statistik" at Bielefeld University. It will take place on Tuesday, 8 July 2025 from 12:00 Uhr - 13:00 in W9-109.

Knowledge Tracing (KT) is the attempt to estimate student learning from observed behavior over time. This is typically performed with latent variable models (especially Hidden Markov Models) where the latent variable is student learning and the observed variable are data acquired from student behavior in e-learning systems, especially their success on exercises. The talk will introduce KT, cover the most prominent classic KT models, shortly discuss current machine learning research into KT, and propose potential new directions for KT research that is more compelling both from a modeling perspective and from the perspective of educational utility.

ReproducibiliTea Meeting

2025-07-01T15:15:48+02:00

“ReproducibiliTea” is an informal meeting of scholars interested in topics such as Open Science, reproducible research, and good scientific practice. It is organized by Katja Politt at the Bielefeld University. It will take place on Friday, 4 July 2025 at 12:00 (noon) in V4-116.

“ReproducibiliTea” is a monthly informal meeting, focused on topics such as Open Science, reproducible research and good scientific practice.

Each session features a group discussion of a selected paper. The discussion can take place in either English or German. The paper for the present discussion is:

"The role of environmental ethics and sustainability in research ethics for linguistics: what is currently available in terms of guidance and how might that be improved?" by Eisenbeiß et al. (2025) (https://doi.org/10.1515/ling-2023-0203 ).

Everybody is welcome to join and discuss, the participation is also possible via Zoom, the Zoom link can be found at https://burst.pages.ub.uni-bielefeld.de/burst/.

Talk on "Estimation of Latent Group Structures in Time-Varying Panel Data Models"

2025-06-24T23:13:51+02:00

"Estimation of Latent Group Structures in Time-Varying Panel Data Models" is the title of Paul Heimerl talk, which is being organised by the "Zentrum of Statistik" at Bielefeld University. It will take place on Tuesday, 24 June 2025 from 12:00 Uhr - 13:00 Uhr in W9-109.

We introduce a panel data model where coefficients vary both over time and the cross-section. Slope coefficients change smoothly over time and follow a latent group structure, being homogeneous within but heterogeneous across groups. The group structure is identified using a pairwise adaptive group fused-Lasso penalty. The trajectories of time-varying coefficients are estimated via polynomial spline functions. We derive the asymptotic distributions of the penalized and post-selection estimators and show their oracle efficiency. A simulation study demonstrates excellent finite sample properties. An application to the pace of microclimate change highlights the relevance of addressing cross-sectional heterogeneity and time-variance in empirical settings.

Talk on "More on Uncertainty in Machine Learning"

2025-06-05T23:18:38+02:00

"More on Uncertainty in Machine Learning" is the title of Prof. Dr. Göran Kauermann talk, which is being organised by the "Zentrum of Statistik" at Bielefeld University. It will take place on Tuesday, 10 June 2025 from 12:00 Uhr - 13:00 Uhr in W9-109.

The quantification of uncertainty is continuously gaining interest in the machine learning community. Tackling this question can be done with statistical tools, getting back to the foundation of statistics, namely how to cope with uncertainty and ignorance. The talk extends our previous work on this topic and presents some new results and insights. We discuss different sources of uncertainty and advocate an increased use of statistics and statistical methods in the machine learning world. To showcase our view we look at ambiguity in computer linguistics, we sketch the use of statistics in network weight reconstruction and focus on labelling ambiguity. While these examples are heterogeneous, the data centric view and hence its statistical foundation serves as common thread.

Talk on "Advances in gradient boosting approaches for geoadditive models"

2025-05-21T13:06:02+02:00

"Advances in gradient boosting approaches for geoadditive models" is the title of Lars Knieper talk, which is being organised by the "Zentrum für Statistik" at Bielefeld University. It will take place on Tuesday, 27 May 2025 from 12:00 to 13:00 in W9-109.

One of the key features of model-based component-wise gradient boosting is a data-driven variable selection mechanism which takes place while estimating the effects simultaneously. When spatial effects, of areal or point-reference data, are included as a potential model-component a drastic increase of chosen fixed effects can be observed. This is accompanied by a high selection frequency of the spatial component without achieving an impactful reduction of the loss function. To address this ineffective variable selection, we propose to eliminate the competition between fixed and spatial effects by separating the spatial part from the component-wise mechanism and ensuring complete estimation in each iteration. Additionally, we suggest using spatial cross-validation, which accounts for the auto-correlated structure of spatial data when constructing folds. Our approach is applied to yield deviation estimates of coffee farmers in Colombia, resulting in a sparser model with improved predictive performance. In addition, there will be an extended outlook on spatial confounding—an estimation bias arising from collinearity between fixed and spatial effects—which tends to be more pronounced in gradient boosting than in maximum likelihood approaches. A straightforward remedy, known as Spatial+, proves effective in mitigating this issue.

Talk on "Insights from cognitive neuroscience – statistical models to understand behavior"

2025-05-07T22:55:58+02:00

"Insights from cognitive neuroscience – statistical models to understand behavior" is the title of Prof. Dr. Christoph Kayser talk, which is being organised by the "Zentrum of Statistik" at Bielefeld University. It will take place on Tuesday, 13 May 2025 from 12:00 Uhr - 13:00 Uhr in W9-109.

In this talk I will present studies on how we (and our brain) combine multisensory signals to guide behavior. The talk will focus on examples such as spatial ventriloquism and multisensory perceptual decision making, but also speak more broadly about how well we can explain behavior in typical laboratory tasks. The goal of the presentation is to illustrate the different types of statistical models employed in these studies, including Bayesian approaches and drift-diffusion (random walk) models and how they can be used to understand behavior, but also how we can link them to neural processes in the brain.

Talk on "When to Collect Data? A Sobol Index Strategy for Model-Based Experimental Design"

2025-04-25T19:54:46+02:00

“When to Collect Data? A Sobol Index Strategy for Model-Based Experimental Design” is the title of Houda Yaqine talk, which is being organised by the "Zentrum für Statistik" at Bielefeld University. It will take place on Tuesday, 29 April 2025 from 12:00 to 13:00 in W9-109.

Optimizing measurement timing in experimental studies is crucial for reducing animal sacrifices, especially when investigating systems characterized by ordinary differential equations (ODEs). Conventional sampling strategies frequently collect excessive data points without substantially improving parameter estimation accuracy. Our research introduces a new experimental design framework that analyzes the temporal evolution of variance-based sensitivities in ODE-driven processes. By integrating time-dependent Sobol indices with underlying system dynamics, this approach identifies the most informative sampling points, particularly during phases where parameter interactions are the primary contributors to system variance. Both theoretical analyses and computational experiments confirm that our strategically selected measurement times yield better parameter estimation outcomes compared to traditional protocols.

Talk on “Using Subspace Methods for the Estimation of Approximate Dynamic Factor Models “

2025-04-11T14:27:10+02:00

“Using Subspace Methods for the Estimation of Approximate Dynamic Factor Models” is the title of Dietmar Bauers's lecture, which is being organised by the "Zentrum für Statistik" at Bielefeld University. It will take place on Tuesday, 15 April 2025 from 12:00 to 13:00 in W9-109.

For multivariate time series with a large number of variables classical vector autoregressive (VAR) models are not appropriate because they contain too many parameters. Alternatively in the literature in such situations factor models are used to reduce the dimensionality. Approximate dynamic factor models represent the high-dimensional time series as generated by a common factor part and idiosyncratic terms, where the common factors are latent. Estimating the dynamics of the common factors often is done using a VAR model for the principal components. The estimation of more flexible state space models via maximum likelihood methods is more complicated. Subspace methods are a numerically stable alternative that can be used in this respect. In this talk we show that the subspace methods provide a very robust and computationally simple means to obtain consistent estimators for the latent factor dynamics.

EuADS Summer School “Automated Data Science” starts in July

2025-03-12T15:14:12+01:00

The EuADS Summer School will take place in Luxembourg from 8th to 11st July 2025. The topic for this year's event is Automated Data Science. Registration for the summer school, which is sponsored and co-organized by BiCDaS, is now open on the homepage of the European Association for Data Science.

more...

Open Science in summer term 2025

2025-03-12T15:12:41+01:00

The seminar "Open Minds for Open Science - An Interdisciplinary Introduction to Open Science" will take place again in the coming summer term 2024. Open Science is becoming more and more important as an interdisciplinary topic, even at the undergraduate level. More and more, Open Science is also in the focus of politics and is politically desired. The seminar is addressed to all students who are interested in Open Science and aspire to a scientific career.

more...

Talk on “Bayesian Penalized Transformation Models for Structured Additive Regression on the Location and Scale of Arbitrary Distributions“

2025-01-09T12:11:34+01:00

“Bayesian Penalized Transformation Models for Structured Additive Regression on the Location and Scale of Arbitrary Distributions” is the title of Johannes Brachem's lecture, which is being organised by the "Zentrum für Statistik" at Bielefeld University. It will take place on Tuesday, 14 January 2025 from 12:00 to 13:00 in W9-109.

Penalized transformation models (PTMs) are a novel form of distribution-free location-scale regression. In PTMs, the shape of the response’s conditional distribution is estimated directly from the data, and structured additive predictors are placed on its location and scale. The core of the model is a monotonically increasing transformation function that relates the response distribution to a reference distribution. The transformation function is equipped with a smoothness prior that regularizes how much the estimated distribution diverges from the reference distribution. These models can be seen as a bridge between conditional transformation models and generalized additive models for location, scale and shape. Markov chain Monte Carlo inference for PTMs can be conducted with the No-U-Turn sampler and offers straightforward uncertainty quantification for the conditional distribution as well as for the covariate effects.

Data Management Plans Workshop

2024-08-03T12:27:28+02:00

The Service Centre for Medical Data Science at Bielefeld University is offering a workshop on data management plans (DMPs). Tailored for researchers in the Medical Faculty and UK OWL, this workshop focuses on crafting DMPs according to DFG criteria. Participants will learn to structure, draft and implement effective DMPs for their research projects.

The workshop will be held online on:

Tuesday, 24 September and 1 October 2024: 16:00-18:00
Thursday, 26 September 2024: 14:00-18:00

Participants are required to prepare an abstract of their research project and create an RDMO account on forschungsdaten.info.

For more information and registration, please visit the registration link and the official page.

Talk on "The Sum Score Model: Specifying and Testing Equally Weighted Composites Using Structural Equation Modeling"

2024-07-14T10:52:48+02:00

"The Sum Score Model: Specifying and Testing Equally Weighted Composites Using Structural Equation Modeling" is the title of Dr. Tamara Schamberger's lecture, which is being organised by the "Zentrum für Statistik" at Bielefeld University. It will take place on Tuesday, 16 July 2024 from 12:00 to 13:00 in W9-109. Dr. Schamberger will present a new method for incorporating sum scores into structural equation modelling (SEM). The approach overcomes current limitations, replicates results of existing methods, and accounts for random measurement errors in the variables forming the sum score.

Talk on "Statistical modelling of narwhal diving behaviour and responses to sound exposure using stochastic differential equations with state-switching coefficients"

2024-06-27T11:00:00+02:00

"Statistical modelling of narwhal diving behaviour and responses to sound exposure using stochastic differential equations with state-switching coefficients" is the title of J.-Prof. Dr. Timo Adam's lecture at the ZeSt Colloquium The lecture of J.-Prof. Dr. Timo Adam from the Faculty of Business Administration and Economics / Department Empirical Methods starts on 02 July at 12:00 in room W9-109.

more...

Talk on "Publication bias: what is it and why does it matter"

2024-06-13T11:00:00+02:00

As part of the ZeSt Colloquium, Dr Alfredo Sánchez-Tójar (associated member of the Collaborative Research Centre "SFB/Transregio 212 A Novel Synthesis of Individualisation across Behaviour, Ecology and Evolution: Niche Choice, Niche Conformity, Niche Construction" (NC³)") from Bielefeld University will give a lecture on "Publication bias: what is it and why does it matter". The lecture starts at 12 noon in room W9-109.

more...

Talk on "Estimating multinomial probit models using the R package "Rprobit""

2024-05-30T11:00:00+02:00

As part of the ZeSt Colloquium, Prof. Dr. Dietmar Bauer from the Faculty of Business Administration and Economics at the University of Bielefeld will give a lecture on 'Estimating multinomial probit models using the R package "Rprobit"' on Tuesday, 4 June 2024. The talk will start at 12 noon in room W9-109.

more...

Talk on "Statistical Contributions in Conflict Research"

2024-05-16T11:00:00+02:00

Prof. Dr. Göran Kauermann from the Institute of Statistics at the Ludwig-Maximilians-Universität in Munich will give a lecture on "Statistical Contributions in Conflict Research" as part of the ZeSt Colloquium. The lecture will take place from 12 to 13 in room W9-109.

more...

Talk on "Fixed-event forecasting systems and efficiency measurement"

2024-05-02T11:00:00+02:00

On Tuesday, the 7th of May, Prof. Dr Haupt from the University of Passau will give a talk on "Fixed-event forecasting systems and efficiency measurement". The talk as part of the ZeSt Colloquium will start at 12 noon in room W9-109. Prof Dr Haupt's talk will be in English.

more...

Talk on "Signal detection of adverse drug reactions: The Bayesian Power generalised Weibull Shape Parameter test"

2024-04-16T13:52:10+02:00

Julia Dyck from Bielefeld University will give a talk on "Signal detection of adverse drug reactions: The Bayesian Power generalised Weibull Shape Parameter test" as part of the ZeSt Colloquium on Tuesday, 23 April 2024. The talk starts at 12 midday in room W9-109.

more...

Computational Social Science Workshop

2024-03-14T11:15:00+01:00

Computational Social Science (CSS) is the topic of a two-day workshop organised by the Working Group on Applied Social Data Science (Kühne) at the Faculty of Sociology at Bielefeld University (ZiF). CSS has established itself as an independent field of research. It examines sociological, political and economic research questions that are often based on large, unstructured amounts of data. Computational methods are used. On 6 and 7 May, the workshop will offer CSS researchers from all over Germany the opportunity to network, exchange experiences and discuss the latest developments. The workshop will focus on the integration of innovative methods and data with social science theory in order to improve the quality and relevance of current CSS research and to contribute to scientific and societal debates. More information on the programme and registration is available at computational-social-science.org.

more...

Second edition "Woche der Forschungskompetenzen"

2024-02-05T11:33:22+01:00

In March, Bielefeld University will host the second "Woche der Forschungskompetenzen". Organised by the Competence Centre for Research Data (a cooperation between Bielefeld University Library and BITS) and the Interdisciplinary Research on Conflict and Violence (IKG), the week from 4 to 8 March is dedicated to skills that are currently given less attention in university curricula. Topics include rights to texts and data, privacy and ethics, and scholarly publishing. The week's lectures, most of which will be held in German, are aimed at all empirical researchers at Bielefeld University, but especially at "young academics".

more...