Zum Inhalt
Fakultät Statistik
Researcher

Dr. Alexander Munteanu

Contact

Office: mathematics building, room E16b
Phone: +49 231 755 - 7885
E-Mail: alexander.munteanu@tu-dortmund.de

Address: Fakultät Statistik
Technische Universität Dortmund
44221 Dortmund

Office hours: by appointment

Alexander Munteanu

About me

After receiving my PhD in theoretical computer science under supervision of Christian Sohler, I am now a researcher in the statistics department at TU Dortmund in the group led by Katja Ickstadt.

I was PI in a project on large-scale and high-dimensional regression problems within the large-scale collaborative research center SFB 876 (completed 2022).

Being a PI in the interdisciplinary research area From Prediction to Agile Interventions in the Social Sciences (FAIR) offers me a great opportunity to transfer and further develop innovative statistical and data science methods for their application in the social sciences.

I am involved in the establishment of the TU Dortmund Center for Data Science & Simulation (DoDaS) as a founding member in the position of managing director.

I am happy to advise my students Simon Omlor (postdoc) and Amer Krivošija (postdoc).

Research interests

I am mainly interested in the design and analysis of algorithms for tackling the challenges of massive data and high dimensionality. I am also interested in collaborating on possible applications. My research involves several scientific fields such as

  • machine learning theory and algorithms,
  • randomized numerical linear algebra,
  • streaming and distributed algorithms,
  • mathematical statistics,
  • computational geometry,
  • convex optimization.

Publications

2024

  • Susanne Frick, Amer Krivošija, Alexander Munteanu.
    Scalable learning of Item Response Theory models.
    International Conference on Artificial Intelligence and Statistics (AISTATS), 2024.

2023

  • Alexander Munteanu, Simon Omlor, David P. Woodruff.
    Almost linear constant-factor sketching for ℓ₁ and logistic regression.
    International Conference on Learning Representations (ICLR), 2023.
     
  • Tung Mai, Alexander Munteanu, Cameron Musco, Anup B. Rao, Chris Schwiegelshohn, David P. Woodruff.
    Optimal sketching bounds for sparse linear regression.
    International Conference on Artificial Intelligence and Statistics (AISTATS), 2023.
     
  • Alexander Munteanu.
    Coresets and sketches for regression problems on data streams and distributed data.
    Machine Learning under Resource Constraints, Volume 1 - Fundamentals, pp. 85-97, 2023.
     
  • Zeyu Ding, Katja Ickstadt, Alexander Munteanu.
    Bayesian analysis for dimensionality and complexity reduction.
    Machine Learning under Resource Constraints, Volume 3 - Applications, pp. 58-70, 2023.

2022

  • Alexander Munteanu, Simon Omlor, Zhao Song, David P. Woodruff.
    Bounding the width of neural networks via coupled initialization - A worst case analysis.
    International Conference on Machine Learning (ICML), 2022.
  • Alexander Munteanu, Simon Omlor, Christian Peters.
    p-Generalized probit regression and scalable maximum likelihood estimation via sketching and coresets.
    International Conference on Artificial Intelligence and Statistics (AISTATS), 2022.

2021

  • Alexander Munteanu, Simon Omlor, David P. Woodruff.
    Oblivious sketching for logistic regression.
    International Conference on Machine Learning (ICML), 2021.

2020

  • Leo N. Geppert, Katja Ickstadt, Alexander Munteanu, Christian Sohler.
    Streaming statistical models via Merge & Reduce.
    International Journal of Data Science and Analytics, 10(4):331-347, 2020.

2019

  • Stefan Meintrup, Alexander Munteanu, Dennis Rohde.
    Random projections and sampling algorithms for clustering of high-dimensional polygonal curves.
    Advances in Neural Information Processing Systems (NeurIPS), 2019.
     
  • Alexander Munteanu, Amin Nayebi, Matthias Poloczek.
    A framework for Bayesian optimization in embedded subspaces.
    International Conference on Machine Learning (ICML), 2019.
     
  • Amer Krivošija, Alexander Munteanu.
    Probabilistic smallest enclosing ball in high dimensions via subgradient sampling.
    Symposium on Computational Geometry (SoCG), 2019.
    European Workshop on Computational Geometry (EuroCG), 2019.

2018

  • Alexander Munteanu, Chris Schwiegelshohn, Christian Sohler, David Woodruff.
    On coresets for logistic regression.
    Advances in Neural Information Processing Systems (NeurIPS), 2018.
     
  • Alexander Munteanu.
    On large-scale probabilistic and statistical data analysis.
    PhD Thesis. TU Dortmund University, 2018.
     
  • Kristian Kersting, Alejandro Molina, Alexander Munteanu.
    Core dependency networks.
    AAAI Conference on Artificial Intelligence (AAAI), 2018.
     
  • Alexander Munteanu, Chris Schwiegelshohn.
    Coresets - methods and history: a theoreticians design pattern for approximation and streaming algorithms.
    KI special issue on "Algorithmic Challenges and Opportunities of Big Data", 32(1):37-53, 2018.

2017

  • Leo N. Geppert, Katja Ickstadt, Alexander Munteanu, Jens Quedenfeld, Christian Sohler.
    Random projections for Bayesian regression.
    Statistics and Computing, 27(1):79-101, 2017.

2016

  • Alexander Munteanu, Max Wornowizki.
    Correcting statistical models via empirical distribution functions.
    Computational Statistics, 31(2):465-495, 2016.

2014

  • Dan Feldman, Alexander Munteanu, Christian Sohler.
    Smallest enclosing ball for probabilistic data.
    Symposium on Computational Geometry (SoCG), 2014.
     
  • Marc Heinrich, Alexander Munteanu, Christian Sohler.
    Asymptotically exact streaming algorithms.
    ArXiv preprint, CoRR abs/1408.1847, 2014.
     

Teaching

My interdisciplinary teaching activities are listed below. Students interested in Bachelor's or Master's theses in either Computer Science, Statistics, or Data Science may contact me via email.