Generative flow-based models in application to uncertainty modeling for machine learning tasks

2021/43/B/ST6/02853

Keywords

Descriptors

  • ST6_011: Uczenie maszynowe, statystyczne przetwarzanie danych i zastosowania w przetwarzaniu sygnałów (np. mowy, obrazów, filmów)

Panel

ST6 – Computer science and informatics: informatics and information systems, computer science, scientific computing, intelligent systems

Host institution

Politechnika Wrocławska

Voivodeship: dolnośląskie

Other projects carried out by the institution »

Principal investigator (from the host institution)

dr hab. Maciej Mateusz Zięba

Number of co-investigators in the project: 3

Call

OPUS 22 – announced on 2021-09-15

Amount awarded
872 800 PLN
Project start date
2022-08-01
Project end date
2026-07-31
Project duration
48 months (the same as in the proposal)
Project status
Pending project

Project description

Download the project description in a pdf file

Note — project descriptions were prepared by the authors of the applications themselves and placed in the system in an unchanged form.

Information in the final report

Publication in academic press/journals (2)

  1. NodeFlow: towards end-to-end flexible probabilistic regression on tabular data

    Authors:
    Patryk E. Wielopolski, Oleksii Furman, Maciej M. Zięba
    Academic press:
    Entropy (year: 2024, volume: Deep Generative Modeling: Theory and Applications, pages: 46038), Publisher: MDPI
    Status:
    Published
    DOI:
    10.3390/e26070593 – link to the publication
  2. Dimma: Semi-supervised low-light image enhancement with adaptive dimming

    Authors:
    Wojciech M. Kozłowski, Michał Szachniewicz, Michał Stypułkowski, Maciej M. Zięba
    Academic press:
    Entropy (year: 2024, volume: vol 26, Deep Generative Modeling: Theory and Applications, pages: 46046), Publisher: MDPI
    Status:
    Published
    DOI:
    10.3390/e26090726 – link to the publication

Articles in post-conference publications (6)

  1. Modeling uncertainty in personalized emotion prediction with normalizing flows

    Authors:
    Piotr D. Miłkowski, Konrad Karanowski, Patryk E. Wielopolski, Jan Kocoń, Przemysław Kazienko, Maciej M. Zięba
    Conference:
    2023 IEEE International Conference on Data Mining Workshops (ICDMW) (year: 2023, volume: 23nd IEEE International Conference on Data Mining Workshops (ICDMW),, pages: 757-766), Publisher: Institute of Electrical and Electronics Engineers
    Date:
    conference 1.12.2023-4.12.2023
    Status:
    Published
    DOI:
    10.1109/ICDMW60847.2023.00103 – link to the publication
  2. RegFlow: Probabilistic Flow-Based Regression for future prediction

    Authors:
    Maciej Zięba, Marcin Przewięźlikowski, Marek Śmieja, Jacek Tabor, Tomasz Trzciński, Przemysław Spurek
    Conference:
    Communications in Computer and Information Science (year: 2024, volume: Asian Conference on Intelligent Information and Database Systems, pages: 267-279), Publisher: Springer
    Date:
    conference 15-18.04.2024
    Status:
    Published
    DOI:
    10.1007/978-981-97-5934-7_23 – link to the publication
  3. Classical out-of-distribution detection methods benchmark in text classification tasks

    Authors:
    Mateusz Baran, Joanna Baran, Mateusz A. Wójcik, Maciej M. Zięba, Adam F. Gonczarek
    Conference:
    Proceedings of the Student Research Workshop (year: 2023, volume: The 61st Annual Meeting of the Association for Computational Linguistics, ACL-SRW 2023, pages: 119-129), Publisher: Association for Computational Linguistics
    Date:
    conference 10-12.07.2023
    Status:
    Published
    DOI:
    10.18653/v1/2023.acl-srw.20 – link to the publication
  4. Probabilistically plausible counterfactual explanations with normalizing flows

    Authors:
    Patryk E. Wielopolski, Oleksii Furman, Jerzy Stefanowski, Maciej M. Zięba
    Conference:
    Frontiers in Artificial Intelligence and Applications. ISSN 0922-6389; vol. 392 (year: 2024, volume: ECAI 2024 : 27th European Conference on Artificial Intelligence,, pages: 954-961), Publisher: IOS Press
    Date:
    conference 19–24.10.2024
    Status:
    Published
    DOI:
    10.3233/FAIA240584 – link to the publication
  5. FlowHMM: fow-based continuous hidden Markov models

    Authors:
    Paweł Lorek, Rafał Nowak, Tomasz Trzciński, Maciej M. Zięba
    Conference:
    Advances in Neural Information Processing Systems. ISSN 1049-5258; 35 (year: 2022, volume: 35th Conference on Neural Information Processing Systems, NeurIPS 2022, pages: 46039), Publisher: Neural Information Processing Systems Foundation
    Date:
    conference 28.11-9.12.2022
    Status:
    Published
    DOI:
    link to the publication
  6. Treeflow: going beyond tree-based parametric probabilistic regression

    Authors:
    Patryk E. Wielopolski, Maciej M. Zięba
    Conference:
    Frontiers in Artificial Intelligence and Applications (year: 2023, volume: ECAI 2023 : 26th European Conference on Artificial Intelligence, pages: 2631-2638), Publisher: IOS Press
    Date:
    conference 30.09–4.10.2023
    Status:
    Published
    DOI:
    10.3233/FAIA230570 – link to the publication

« New search

Graphic design and implementation: Jakub Piotrowski