David Rügamer

About

I am currently Interim Professor for Data Science at the LMU Munich. Before I was a Lecturer at the Chair of Statistical Learning and Data Science in the Department of Statistics at the LMU Munich. I have a Bachelor's Degree (B.Sc.) in Statistics with a minor in Computer Science, a Master's Degree (M.Sc.) in Statistics with specialization in theory and a Ph.D. (Dr.rer.nat.) in Statistics with focus on functional data analysis, gradient boosting and statistical inference. During my Ph.D. I worked for the Biostatistics Working Group at the LMU, where I did my Ph.D. from Oct 2014 to Jun 2018 under the supervision of Prof. Dr. Sonja Greven. My work was partly funded by the Emmy Noether project ‘Statistical Methods for Longitudinal Functional Data’.
After a short PostDoc stay at the same group I worked for 1 year as a Senior Data Scientist in the industry with strong focus on Data Engineering and Deep Learning research. Together with other Postdocs in our group I am currently leading the subgroups Machine Learning and Deep Learning.

You can find more information about me on my personal website.

Contact

Institut für Statistik

Ludwig-Maximilians-Universität München

Ludwigstraße 33

D-80539 München

David.Ruegamer [at] stat.uni-muenchen.de

Teaching

Summer 2021:
- Applied Deep Learning with TensorFlow and PyTorch
- Predictive Modeling
Winter 2020/21:
- Deep Learning
Summer 2020:
- Introduction to R
- Predictive Modeling

Thesis

I offer various topics for Master theses mainly but not exclusively on the topics listed below. If you are interested, please send me your field of interest or your ideas, a CV and your current transcript of records.

Research Interests

My main research interest lies in the combination of structured additive models and deep neural networks. In deep learning, combinations of additive models and deep neural networks are sometimes referred to as wide and deep learning, although there is no formal definition and the types of networks do not necessarily focus on structured models for the wide part. Semi-structured deep learning is my very own interpretation that has several important properties (read the paper :)). The broader topic I am interested in is Uncertainty quantification in deep learning. Other topics I work on / find interesting:

Selective Inference
Gradient Boosting
Distributional Regression
Transformation Models / Normalizing Flows
Functional Data Analysis
Time Series Analysis
Survival Analysis
Recommender Systems, Matrix and Tensor Factorization
Optimization

News

News on my personal homepage

You Can Find me on

Software

deepregression Semi-Structured Deep Distributional Regression on Github
selfmade SELective inference For Mixed and ADditive model Estimators on Github
FDboost Boosting Functional Regression Models on CRAN and github
cAIC4 Conditional Akaike Information Criterion for ‘lme4’ on CRAN and github
iboost Inference for Model-based Boosting on github
coinflibs Conditional Inference after Likelihood-based Selection on github
effortless efficient operations on row-wise tensor product linked evaluations with special structures on github

References

*Coors S, *Schalk D, Bischl B, Rügamer D (2021) Automatic Componentwise Boosting: An Interpretable AutoML System. ECML-PKDD Workshop on Automating Data Science.
Berninger C, Stöcker A, Rügamer D (2021) A Bayesian Time-Varying Autoregressive Model for Improved Short- and Long-Term Prediction. Accepted at the Journal of Forecasting.
link|pdf.
Baumann PFM, Hothorn T, Rügamer D (2021) Deep Conditional Transformation Models. Accepted at ECML-PKDD 2021.
link|pdf.
Kopper P, Pölsterl S, Wachinger C, Bischl B, Bender A, Rügamer D (2021) Semi-Structured Deep Piecewise Exponential Models. In: In: Greiner R , In: Kumar N , In: Gerds TA , In: Schaar M van der (eds) Proceedings of AAAI Spring Symposium on Survival Prediction - Algorithms, Challenges, and Applications 2021, pp. 40–53. PMLR.
link|pdf.
Rügamer D, Shen R, Bukas C et al. (2021) deepregression: a Flexible Neural Network Framework for Semi-Structured Deep Distributional Regression. arXiv:2104.02705 [stat].
link|pdf.
Liew B, Lee HY, Rügamer D et al. (2021) A novel metric of reliability in pressure pain threshold measurement. Scientific Reports (Nature).
Bender A, Rügamer D, Scheipl F, Bischl B (2021) A General Machine Learning Framework for Survival Analysis. In: In: Hutter F , In: Kersting K , In: Lijffijt J , In: Valera I (eds) Machine Learning and Knowledge Discovery in Databases, pp. 158–173. Springer International Publishing.
link | pdf.
Goschenhofer J, Hvingelby R, Rügamer D, Thomas J, Wagner M, Bischl B (2021) Deep Semi-Supervised Learning for Time Series Classification. arXiv:2102.03622 [cs].
link|pdf.
Fritz C, Dorigatti E, Rügamer D (2021) Combining Graph Neural Networks and Spatio-temporal Disease Models to Predict COVID-19 Cases in Germany. arXiv:2101.00661 [cs, stat].
link|pdf.
Rügamer D, Pfisterer F, Bischl B (2020) Neural Mixture Distributional Regression. arXiv:2010.06889 [cs, stat].
link|pdf.
Rügamer D, Kolb C, Klein N (2020) A Unified Network Architecture for Semi-Structured Deep Distributional Regression. arXiv:2002.05777 [cs, stat].
link|pdf.
Rügamer D, Greven S (2020) Inference for L2-Boosting. Statistics and Computing 30, 279–289.
link|pdf.
Liew BXW, Rügamer D, Stöcker A, De Nunzio AM (2020) Classifying neck pain status using scalar and functional biomechanical variables – development of a method using functional data boosting. Gait & Posture 75, 146–150.
link.
Liew B, Rügamer D, De Nunzio A, Falla D (2020) Interpretable machine learning models for classifying low back pain status using functional physiological variables. European Spine Journal 29, 1845–1859.
link.
Liew BXW, Rügamer D, Abichandani D, De Nunzio AM (2020) Classifying individuals with and without patellofemoral pain syndrome using ground force profiles – Development of a method using functional data boosting. Gait & Posture 80, 90–95.
link.
Liew BXW, Peolsson A, Rügamer D et al. (2020) Clinical predictive modelling of post-surgical recovery in individuals with cervical radiculopathy – a machine learning approach. Scientific Reports.
link.
Brockhaus S, Rügamer D, Greven S (2020) Boosting Functional Regression Models with FDboost. Journal of Statistical Software 94, 1–50.
Rügamer D, Baumann P, Greven S (2020) Selective Inference for Additive and Mixed Models. arXiv preprint arXiv:2007.07930.
link|pdf.
Rügamer D, Brockhaus S, Gentsch K, Scherer K, Greven S (2018) Boosting factor-specific functional historical models for the detection of synchronization in bioelectrical signals. Journal of the Royal Statistical Society: Series C (Applied Statistics) 67, 621–642.
Rügamer D, Greven S (2018) Selective inference after likelihood-or test-based model selection in linear models. Statistics & Probability Letters 140, 7–12.
Säfken B, Rügamer D, Kneib T, Greven S (2018) Conditional model selection in mixed-effects models with caic4. to appear in the Journal of Statistical Software.
Brockhaus S, Rügamer D, Greven S (2017) Boosting Functional Regression Models with FDboost. to appear in the Journal of Statistical Software.
Klüser L, Holler PJ, Simak J et al. (2016) Predictors of sudden cardiac death in Doberman Pinschers with dilated cardiomyopathy. Journal of veterinary internal medicine 30, 722–732.
Gillhuber J, Rügamer D, Pfister K, Scheuerle MC (2014) Giardiosis and other enteropathogenic infections: a study on diarrhoeic calves in Southern Germany. BMC research notes 7, 112.