Machine learning approaches in the modeling of hydrological extremes

• Thesis title in Czech: Modelování hydrologických extrémů s využitím strojového učení
• Thesis title in English: Machine learning approaches in the modeling of hydrological extremes
• Key words: modelování; strojové učení; neuronové sítě; hydrologie; extrémy
• English key words: modeling; machine learning; neural networks; hydrology; extremes
• Academic year of topic announcement: 2021/2022
• Thesis type: dissertation
• Thesis language: angličtina
• Department: Department of Physical Geography and Geoecology (31-330)
• Supervisor: prof. RNDr. Jakub Langhammer, Ph.D.
• Author: hidden - assigned by the advisor
• Date of registration: 05.10.2021
• Date of assignment: 06.10.2021

Preliminary scope of work

Modelování hydrologických extrémů s využitím strojového učení

Preliminary scope of work in English

The PhD project is focused on the applications of machine learning (ML) and deep learning (DL) models for the analysis of hydrological extremes in montane basins.
The selected ML and DL techniques are used for the analysis of conditionality and links between causal factors, changes in frequency, seasonality, magnitude or for the search for regularities and possibilities of prediction of hydrological extreme phenomena, including both floods and droughts.
The principal modeling techniques, used for this thesis include neural networks, support vector machines, and Deep Learning models.
The study area presents the selected stations of the headwaters of montane streams in different physiographic conditions with long observation time series, supplemented by experimental high-frequency monitoring at the stations, operated by the Department of Physical Geography and Geoecology.