Presentation #541.15 in the session “Computational Augmentation to Surveys and Science Programs”.
Morphological classification of galaxies yields important information in the understanding of the evolution of the universe. To-date, morphological classification is primarily done visually by volunteer scientists through programs such as Galaxy Zoo. To reduce the reliance on volunteer scientists, increase accuracy, and efficiently classify the millions of sources detected by sky surveys, a new technique is presented based on machine learning and the CyMorph Python package to automatically classify galaxies. With CyMorph, parameters such as smoothness (S), asymmetry (A), gradient pattern analysis (GPA), and Shannon Entropy (H) were calculated for a training set of pre-classified galaxies and used as a training set for a machine learning algorithm created with scikit-learn. After completing the training and testing of the machine learning algorithm, the overall accuracy was calculated as 93.4%. The algorithm will be deployed to automatically classify supernova host galaxies detected by the Carnegie Supernova Project and various sky surveys and aid in the understanding of the evolution of the universe.