Discovery of astronomical objects in galaxies by means of deep learning

Author:

Haindavi Ponguvala,Kumar Sunil,Ganesh ,Anil Kumar R.,Pushkarna Gaurav

Abstract

The study of space exploration has been studied for a very long time, and as technology has developed, so too have the methods and techniques employed, along with the quantity and type of data acquired. We now receive so much astronomical data, and so much brand new data is being generated every day, that it is physically impracticable to examine it all only by human work. In our study, we look at a number of astronomers face while working with this massive amount of data, and they use deep learning techniques to discover the best data for each objective. [1]previously SVM, KNN, the random forest approach, decision trees, and other multi-class classification algorithms are all used in the methodology. Galaxies' propensity to belong to specific classes is forecasted using regression. even if the findings from the random forest method were the best it was unable to effectively divide galaxies into the five groups. This approach does not explain real-time categorization and does not take outliers into consideration. The model's adaptability is constrained. This categorization scheme is unable to account for the modelling of galaxies as well as their evolution. Here, we suggest using Inception v3 for categorization and VGG-19 for image analysis. Segmentation is a method for discovering and classifying galaxies. These techniques greatly advance certain fields of study where there are enormous volumes of duplicate data that must be deleted in accordance with the demands of the study, thanks to Python's high performance in the investigation of image processing and computer vision.. As a result, there is less of a need for researchers to carefully sort through all of the data that has been collected from satellite telescopes, sky surveys, etc. [2]

Publisher

EDP Sciences

Reference16 articles.

1. Dere Shivani, Fatima Maziya, Jagtap Rutuja and Shardoor Nikhilkumar B “Anomaly Detection in Astronomical Objects of Galaxies Using Deep Learning,”7th International Conference on Advanced Computing & Communication Systems (ICACCS),(2021)

2. Krizhevsky A., Sutskever I., and Hinton G. E., “Imagenetclassification with deep convolutional neuralnetworks,” in Advances in Neural Information Processing Systems, (2012), pp. 1097–1105.

3. Gidaris S., Singh P., and Komodakis N., “Unsupervised Representation Learning by Predicting Image Rotations,” in International Conference on Learning Representations, (2018).

4. Caron M., Bojanowski P., Joulin A., and Douze M., “Deep Clustering for Unsupervised Learning of Visual Features,” in European Conference on Computer Vision, (2018).

5. Doersch C., Gupta A., and Efros A. A., “Unsupervised Visual Representation Learning by Context Prediction,” (2015).

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3