Project detail

Quality internal grants at BUT

Duration: 01.02.2021 — 31.01.2023

Funding resources

Evropská unie - Interní grantová soutěž

- whole funder (2021-02-01 - 2023-01-31)

On the project

The problem of the distance metric for high-dimensional spectra is addressed. Distance (or similarity) measurement is a key component of numerous machine learning algorithms. By default, the Euclidean distance is dominantly used, which is known as a poor metric for high-dimensional data. Overlooking the importance of metric selection leads to counterintuitive results and limited performance. By implementing alternative metrics, a rapid boost of performance and model interpretability is expected with a high potential to influence the community.

Description in English
The aim of the project is to create a competition for student research grants and its pilot verification. The creation of a new competition will contribute to the development of cross-sectional skills of doctoral students, and thus acquire competencies for work in science and research in the future and increase their success in submitting scientific projects to national and international competitions.

Mark

CEITEC-K-21-6978

Default language

Czech

People responsible

Vrábel Jakub, Ing. - principal person responsible

Units

Advanced instrumentation and methods for material characterization
- (2021-02-01 - 2023-01-31)
Central European Institute of Technology BUT
- (2021-02-01 - 2023-01-31)

Results

VRÁBEL, J.; KÉPEŠ, E.; POŘÍZKA, P.; KAISER, J. Physics-informed ML models for processing of spectroscopic data. 2021.
Detail

VRÁBEL, J.; KÉPEŠ, E.; POŘÍZKA, P.; KAISER, J. Artificial neural network weights penalization and initialization for spectroscopic data. 2021.
Detail

VRÁBEL, J. Physics-informed ML models for processing of spectroscopic data. 2021.
Detail

VRÁBEL, J.; KÉPEŠ, E.; NEDĚLNÍK, P.; BUDAY, J.; CEMPÍREK, J.; POŘÍZKA, P.; KAISER, J. Spectral library transfer between distinct Laser-Induced Breakdown Spectroscopy systems trained on simultaneous measurements. Journal of Analytical Atomic Spectrometry, 2023, vol. 38, no. 4, p. 841-853. ISSN: 1364-5544.
Detail

VRÁBEL, J.; KÉPEŠ, E.; NEDĚLNÍK, P.; POŘÍZKA, P.; KAISER, J. Spectra transfer between distinct LIBS systems using shared standards and machine learning. 2022.
Detail

VRÁBEL, J.; KÉPEŠ, E.; POŘÍZKA, P.; KAISER, J. Distance of Spectroscopic Data. 2022.
Detail

VRÁBEL, J.; KÉPEŠ, E.; POŘÍZKA, P.; KAISER, J. Towards interpretability of ANNs for spectroscopic data: inductive bias, lottery tickets, and input optimization. 2022.
Detail

VRÁBEL, J.; KÉPEŠ, E.; POŘÍZKA, P.; KAISER, J. Artificial Neural Networks for Classification. In Chemometrics and Numerical Methods in LIBS. 1. 2022. p. 213-240. ISBN: 978-1-119-75958-4.
Detail

Responsibility: Vrábel Jakub, Ing.