Machine Learning for the Estimation of Galaxies
Share this Session:
  Pragyansmita Nayak   Pragyansmita Nayak
Technical and Data Architect
CGI Federal Inc.


Wednesday, February 1, 2017
08:30 AM - 09:15 AM

Level:  Technical - Intermediate

This talk looks at "photomorphic" redshift estimation of galaxies. Spectroscopic measurements and color information (photometric) of the galaxies have been used as redshift predictors. This new approach additionally leverages the galaxy's morphology information. Astronomers explore the cosmos not just by observing through the tiny visible window used by our eyes, but also by exploiting the entire electromagnetic spectrum, from radio waves with wavelengths larger than a house to gamma rays with wavelengths 1,000 times smaller than a proton. This has generated a treasure trove of terabytes of open data related to large-scale wide and deep, both ground and space-based, surveys of the different regions of the sky at multiple wavelengths. Machine Learning (ML) is increasingly being employed to develop robust and scalable algorithms and tools to analyze and mine these images and their associated catalogs to determine patterns and trends and solve science problems of this nature. Bayesian Network and GLM algorithms have been applied with promising results. Our view of the universe is closely tied to our understanding of galaxy formation. This work is a step in that direction using the available "smart data" using "intelligent techniques".

Dr. Pragyansmita Nayak is Technical and Data Architect in the Momentum Product Development Group at CGI Federal Inc. She has over 18+ yrs of experience and holds a Ph.D. in Computational Sciences and Informatics (under the esteemed guidance of Prof. Kirk Borne at GMU, Fairfax, VA) and a Bachelors in Computer Science (BITS Pilani, India). Her Ph.D. thesis focused on the application of Machine Learning techniques (Generalized Linear Model and Bayesian Networks) on the estimation of the Photomorphic Redshift of the galaxies. She previously worked with Novell India, Wipro Global R&D, and IIT Madras ERNET PoP. She is an avid meetup and hackathon participant. She won the 2016 AngelHack DC HPE Haven OnDemand(HOD) challenge. She won the best graduate poster prize at the 2015 Mid-Atlantic Sigma Xi competition. She receives rave reviews on her presentations involving data science and technical architecture and is a Dataversity presenter.?

Close Window