COSMOS2025: A Machine Learning Census of Massive Quiescent Galaxies at 2.5 < z < 5

This repository contains: 🚀 Pre-trained ML model for identifying massive high-redshift quiescent galaxies in COSMOS2025 catalog. 📊 Full classified catalog (machine-readable tables). 📜 Documentation for reproducibility. Description: 1. Trained Models - ML_Model_Classifier.pkl - ML model classifer - ML_Model_Scaler.pkl - ML model scaler fitted on training data 2. Classified Catalog - COSMOS2025_ml_classifications.csv- Columns: 'ra', 'dec', 'hst-f814w', 'uvista-y', 'uvista-j', 'uvista-h', 'uvista-ks', 'irac-ch1', 'irac-ch2', 'nircam_f115w', 'nircam_f150w', 'nircam_f277w', 'nircam_f444w', 'zpdf_med', 'mass_med', 'ssfr_med', 'ML-class', 'SED-class', 'ML-class_prob'.- Column descriptions: COLUMN_DESCRIPTIONS.ipynb 3. Usage (Reproduce results)- File: Quiescent_Galaxy_Classifier_Usage_Example.ipynb- Needed Libraries: Pandas and joblib4. Pre-processing- File: Preprocessing.ipynb- Needed Libraries: Pandas, NumPy, Sklearn and missforest- Needed Data: COSMOS2025_with_missing_values.csv & SAM_Mags.csv