The Role of AGN in Quenching Massive and Ultra-Massive Galaxies: Insights from Machine Learning, Emission Line Diagnostics, and JWST Observations

<p dir="ltr">This thesis investigates AGN-driven quenching in massive galaxies at redshift 3 < z < 4. Analyzing spectroscopic data from ground-based observatories and JWST, we found a high AGN prevalence (60-70%) in ultra-massive systems. We developed a novel "BPT+S" diagram to improve AGN classification and explored machine learning techniques to study galaxy evolution. Comparisons between observed star formation histories and the TNG300 simulation confirmed that AGN feedback is critical for rapid quenching. Our work provides direct observational evidence for the dominant role of AGN in shutting down star formation in the early universe.</p>