Abstract: 4-Month seminar for a PhD project about the development of a classification pipeline for the 4-Metre Multi-Object Spectroscopic Telescope (4MOST), expected to start operation in Q2 2024. The project focuses on the development, implementation and testing of a Bayesian Neural Network for broad and fine probabilistic classification of astronomical sources (stars, galaxies, quasars, etc.) based on the observed high-resolution spectra, using only empirical training data.