The increased volume of images and galaxies surveyed by recent and upcoming projects consolidates the need for accurate and scalable automated AI-driven classification methods. This paper proposes a new algorithm based on a custom neural network architecture for classifying galaxies from deep space surveys. The convolutional neural network (CNN) presented is trained using 10,000 galaxy images obtained from the Galaxy Zoo 2 dataset. It is designed to categorize galaxies into five distinct classes: completely round smooth, in-between smooth (falling between completely round and cigar-shaped), cigar-shaped smooth, edge-on, and spiral. The performance of the proposed CNN is assessed using a set of metrics such as accuracy, precision, recall, F1 score, and area under the curve. We compare our solution with well-known architectures like ResNet-50, DenseNet, EfficientNet, Inception, MobileNet, and one proposed model for galaxy classification found in the recent literature. The results show an accuracy rate of 96.83%, outperforming existing algorithms.

"The potential threat of Near Earth Objects (NEO) requires a constant survey of the night sky to discover potentially dangerous objects and assess their future impact odds. Several ongoing surveys relying on human operators or automated techniques exist. One such example is the EURONEAR blink mini-survey project which over time developed from a pure manual approach to detecting asteroids to semi-automatic methods (NEARBY) using image processing and service-oriented approaches. In this paper, we propose an extension of NEARBY based on an ensemble model comprising three state-of-art machine learning models, some used in similar approaches. The proposed model is designed for a binary classification problem where candidate images may contain an asteroid in their center. Validation on a real-life dataset comprising 11,000 images shows that our ensemble model is capable of recovering about 55% of the asteroids missed by the previous NEARBY automated process while at the same time having a 0.88 recall on the asteroids already detected by NEARBY. Used together with NEARBY our model increased the detection rate from 89% to 95%."

This study explores feature selection for classifying galaxy morphology using the extensive Galaxy Zoo 2 dataset. We investigate supervised and unsupervised learning methods to group galaxies based on key features, aiming to replicate supervised learning results. We evaluate various feature selection methods and compare them to an existing classification approach. Our results demonstrate that a reduced set of features based on adjusted vote fractions improves classification accuracy and potentially reduces computational complexity. While unsuper-vised clustering partially groups galaxies by morphology, further optimization is required. This work suggests that feature selection and unsupervised learning are promising techniques for the efficient classification of large galaxy datasets in upcoming astronomical surveys.