The representation of high-dimensional data in a low dimensional feature space is a cornerstone of data analysis. Nevertheless, feature spaces are rarely investigated in detail, e.g., it is generally unknown how changes of the measurement parameters affect the data’s representation in a feature space. Consequently, the project will evaluate the most common and some of the state-of-the-art feature spaces, i.e., data transformation techniques, to identify the ones which could significantly boost the performance of spectroscopic analysis models.

CEITEC VUT-J-20-6482

Czech

Képeš Erik, Ing., Ph.D. - principal person responsible
Kaiser Jozef, prof. Ing., Ph.D. - fellow researcher

Advanced instrumentation and methods for material characterization
- responsible department (5.3.2020 - not assigned)
Advanced instrumentation and methods for material characterization
- responsible department (1.1.2020 - 31.12.2020)
Central European Institute of Technology BUT
- beneficiary (1.1.2020 - 31.12.2020)

KÉPEŠ, E.; VRÁBEL, J.; POŘÍZKA, P.; KAISER, J. Addressing the sparsity of laser-induced breakdown spectroscopy data with randomized sparse principal component analysis. Journal of Analytical Atomic Spectrometry, 2021, vol. 36, no. 6, p. 1410-1421. ISSN: 1364-5544.
Detail

KÉPEŠ, E.; GORNUSHKIN, I.; POŘÍZKA, P.; KAISER, J. Spatiotemporal spectroscopic characterization of plasmas induced by non-orthogonal laser ablation. ANALYST, 2020, vol. 146, no. 3, p. 920-929. ISSN: 0003-2654.
Detail

Responsibility: Képeš Erik, Ing., Ph.D.