Michael Hagmann

Michael Hagmann

Michael Hagmann, Dr. phil.

I am a postdoctoral researcher at the Prof. Riezler’s StatNLP group. My current research focus is the systematic generation of synthetic time series data to improve algorithm perfromance and privacy preservation. I am also interested in the the application and adapation of empirical methods for the analysis of machine learning experiments.

Research Interests

  • Synthetic Data Generation
  • Differential Privacy
  • Time Series Modelling
  • Empirical Methods in ML
  • Convex and Non-Convex Optimization
  • Bayesian Statistics

News

  • 2023 MAY: Present a poster on inferential reproducibilty @ICLR_2023
  • 2022 SEP: Copresent a tutorial on empirical methods for ML @ECML_2022
  • 2022 JUL: Copresent a tutorial on empirical methods for ML @ICML_2022

Teaching

WS 2023 Formale Grundlagen der Computerlinguistik:
Mathematische Grundlagen		 (material posted on Moodle)
WS 2022 Formale Grundlagen der Computerlinguistik:
Mathematische Grundlagen		 (material posted on Moodle)

Curriculum Vitae

Education

2023 Doctorate in Computational Linguistics, Heidelberg University
2016 Master in Statistics, University of Vienna
2016 Bachelor in Psychology, University of Vienna
2013 Bachelor in Statistics, University of Vienna

Employment History

10/23 Postdoctoral Researcher
StatNLP group, Department for Computational Linguistics, Heidelberg University
09/19 - 09/23 Doctoral Student
StatNLP group, Department for Computational Linguistics, Heidelberg University
10/18 - 08/19 Research Associate
Heinrich-Lanz-Zentrum, Medical Faculty Mannheim, Heidelberg University
04/16 - 09/18 Research Associate
Department for Medical Statistics, Medical Faculty Mannheim, Heidelberg University
06/12 - 03/16 Consultant Statistician
Section for Medical Statistics, CeMSIIS, Medical University of Vienna

Awards

  • 2017 Award for the best Master thesis in applied statistics by the Austrian Statistical Society (ÖSG) pdf

Selected Publications (full list)

  1. Michael Hagmann, Michael Staniek and Stefan Riezler
    Compositionality in Time Series: A Proof of Concept using Symbolic Dynamics and Compositional Data Augmentation
    Transactions on Machine Learning Research, 2025
    pdf | bib 
    @article{hagmann2025compositionality,
  title = {Compositionality in Time Series: A Proof of Concept using Symbolic Dynamics and Compositional Data Augmentation},
  author = {Hagmann, Michael and Staniek, Michael and Riezler, Stefan},
  journal = {Transactions on Machine Learning Research},
  issn = {2835-8856},
  year = {2025},
  url = {https://openreview.net/forum?id=msI02LXVJX}
}
  2. Michael Staniek, Marius Fracarolli, Michael Hagmann and Stefan Riezler
    Early Prediction of Causes (not Effects) in Healthcare by Long-Term Clinical Time Series Forecasting
    Proceedings of Machine Learning Research, 252, 1–30, 2024
    link | bib 
    @inproceedings{staniek2024earlypredictioncausesnot,
  title = {Early Prediction of Causes (not Effects) in Healthcare by Long-Term Clinical Time Series Forecasting},
  author = {Staniek, Michael and Fracarolli, Marius and Hagmann, Michael and Riezler, Stefan},
  year = {2024},
  journal = {Proceedings of Machine Learning Research},
  volume = {252},
  pages = {1--30},
  url = {https://proceedings.mlr.press/v252/staniek24a.html}
}
  3. Stefan Riezler and Michael Hagmann
    Validity, Reliability, and Significance: Empirical Methods for NLP and Data Science - Second Edition
    Synthesis Lectures on Human Language Technologies, Springer, 2024
    link | bib 
    @book{riezler2024,
  author = {Riezler, Stefan and Hagmann, Michael},
  title = {Validity, Reliability, and Significance: Empirical Methods for NLP and Data Science - Second Edition},
  publisher = {Springer},
  series = {Synthesis Lectures on Human Language Technologies},
  editor = {Hirst, Graeme},
  year = {2024},
  isbn = {978-3-031-57064-3},
  doi = {https://doi.org/10.1007/978-3-031-57065-0}
  url = {https://doi.org/10.1007/978-3-031-57065-0}
}
  4. Michael Hagmann, Shigehiko Schamoni and Stefan Riezler
    Validity problems in clinical machine learning by indirect data labeling using consensus definitions
    Machine Learning for Health Symposium (ML4H), ML4H, New Orleans, United States, 2023
    pdf | bib 
    @inproceedings{hagmann2023validity,
  title = {Validity problems in clinical machine learning by indirect data labeling using consensus definitions},
  author = {Hagmann, Michael and Schamoni, Shigehiko and Riezler, Stefan},
  year = {2023},
  month = dec,
  journal = {Machine Learning for Health Symposium},
  journal-abbrev = {ML4H},
  organization = {ML4H},
  publisher = {ML4H},
  city = {New Orleans},
  country = {United States},
  url = {https://arxiv.org/abs/2311.03037}
}
  5. Michael Hagmann, Philipp Meier and Stefan Riezler
    Towards Inferential Reproducibility of Machine Learning Research
    The Eleventh International Conference on Learning Representations (ICLR), Kigali, Rwanda, 2023
    pdf | bib 
    @inproceedings{hagmann2023towards,
  author = {Hagmann, Michael and Meier, Philipp and Riezler, Stefan},
  title = {Towards Inferential Reproducibility of Machine Learning Research},
  journal = {The Eleventh International Conference on Learning Representations},
  journal-abbrev = {ICLR},
  year = {2023},
  city = {Kigali},
  country = {Rwanda},
  url = {https://arxiv.org/abs/2302.04054}
}
  6. Shigehiko Schamoni, Michael Hagmann and Stefan Riezler
    Ensembling Neural Networks for Improved Prediction and Privacy in Early Diagnosis of Sepsis
    Proceedings of Machine Learning Research, 182, PMLR, Durham, NC, USA, 2022
    pdf | bib 
    @inproceedings{schamoni2022,
  author = {Schamoni, Shigehiko and Hagmann, Michael and Riezler, Stefan},
  title = {Ensembling Neural Networks for Improved Prediction and Privacy in Early Diagnosis of Sepsis},
  booktitle = {Proceedings of the 6th Machine Learning for Healthcare Conference},
  year = {2022},
  city = {Durham, NC},
  country = {USA},
  volume = {182},
  series = {Proceedings of Machine Learning Research},
  month = {05--06 Aug},
  publisher = {PMLR},
  url = {https://proceedings.mlr.press/v182/schamoni22a/schamoni22a.pdf}
}
  7. H. A. Lindner, S. Schamoni, T. Kirschning, C. Worm, B. Hahn, F. S. Centner, J. J. Schoettler, M. Hagmann, J. Krebs, D. Mangold, S. Nitsch, S. Riezler, M. Thiel and V. Schneider-Lindner
    Ground truth labels challenge the validity of sepsis consensus definitions in critical illness
    Journal of Translational Medicine, 20(6), 27, 2022
    pdf | bib 
    @article{lindner2022,
  author = {Lindner, H. A. and Schamoni, S. and Kirschning, T. and Worm, C. and Hahn, B. and Centner, F. S. and Schoettler, J. J. and Hagmann, M. and Krebs, J. and Mangold, D. and Nitsch, S. and Riezler, S. and Thiel, M. and Schneider-Lindner, V.},
  title = {Ground truth labels challenge the validity of sepsis consensus definitions in critical illness},
  journal = {Journal of Translational Medicine},
  year = {2022},
  volume = {20},
  number = {6},
  pages = {27},
  doi = {10.1186/s12967-022-03228-7},
  url = {https://doi.org/10.1186/s12967-022-03228-7}
}
  8. Stefan Riezler and Michael Hagmann
    Validity, Reliability, and Significance: Empirical Methods for NLP and Data Science
    Synthesis Lectures on Human Language Technologies, Springer Cham, 2022
    link | bib 
    @book{riezler2022,
  author = {Riezler, Stefan and Hagmann, Michael},
  title = {Validity, Reliability, and Significance: Empirical Methods for NLP and Data Science},
  publisher = {Springer Cham},
  series = {Synthesis Lectures on Human Language Technologies},
  editor = {Hirst, Graeme},
  year = {2022},
  isbn = {9783031010552},
  doi = {https://doi.org/10.1007/978-3-031-02183-1}
  url = {https://doi.org/10.1007/978-3-031-02183-1}
}