Projects funded by the NCN

Information on the principal investigator and host institution

Information of the project and the call

Keywords

Equipment

Delete all

Nowe metody uczenia reprezentacji w Deep Learningu

2017/24/T/ST6/00487

Keywords:

Descriptors:

  • ST6_7: Artificial intelligence, intelligent systems, multi-agent systems
  • ST6_11: Machine learning, statistical data processing and applications using signal processing (e.g. speech, image, video)

Panel:

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

Host institution :

Uniwersytet Jagielloński, Wydział Matematyki i Informatyki

woj. małopolskie

Other projects carried out by the institution 

Principal investigator (from the host institution):

Stanisław Jastrzebski 

Number of co-investigators in the project: 2

Call: ETIUDA 5 - announced on 2016-12-15

Amount awarded: 63 481 PLN

Project start date (Y-m-d): 2017-10-02

Project end date (Y-m-d): 2018-04-01

Project duration:: 6 months (the same as in the proposal)

Project status: Project settled

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

  • Articles in post-conference publications (3)
  1. On the Relation Between the Sharpest Directions of DNN Loss and the SGD Step Length
    Authors:
    S. Jastrzębski, Z. Kenton, N. Ballas, A. Fischer, Y. Bengio, A. Storkey
    Conference:
    International Conference on Learning Representations 2019 (rok: 2019, ), Wydawca: N/A
    Data:
    konferencja 06.05.2019-09.05.2019
    Status:
    Published
    DOI:
    link to the publication
  2. Width of Minima Reached by Stochastic Gradient Descent is Influenced by Learning Rate to Batch Size Ratio
    Authors:
    S. Jastrzębski, Z. Kenton, D. Arpit, N. Ballas, A. Fischer, Y. Bengio, A. Storkey
    Conference:
    Artificial Neural Networks and Machine Learning – ICANN 2018 (rok: 2018, ), Wydawca: Springer International Publishing
    Data:
    konferencja 4-7.10.2018
    Status:
    Published
  3. Dynamical Isometry is Achieved in Residual Networks in a Universal Way for any Activation Function
    Authors:
    W. Tarnowski, P. Warchoł, S. Jastrzębski, J. Tabor, M. A. Nowak
    Conference:
    The 22nd International Conference on Artificial Intelligence and Statistics (rok: 2019, ), Wydawca: PMLR
    Data:
    konferencja 16-18.04.2019
    Status:
    Published
    DOI:
    link to the publication

O Narodowym Centrum Nauki

Finansowanie nauki

Współpraca zagraniczna

Centrum prasowe

Deklaracja dostępności
Przetwarzanie danych osobowych
Polityka cookies

Partnerzy