Lightning talk (L7) Amanda Ibsen (University of Edinburgh)
Theme: Multi-wavelength astronomy
(P3.10) Prompt detection of super-luminous supernovae with deep learning
Deep Learning is an area of Machine Learning that has gained great popularity in the last decade due to its success analyzing both images and sequential data. This makes it an ideal and powerful tool for time domain astronomy and specifically for the challenge of automatic transient detection and classification, that will be a relevant issue with the upcoming LSST data. Using simulated photometry, we attempt to perform light curve automatic classification as early as possible, without making use of any additional features (such as host galaxy photometric resdshift or handcrafted statistical features) other than the time series itself and place special importance in the correct detection of rare events such as Super Luminous Super Novae.
Poster Abstract
P3.10 Amanda Ibsen (University of Edinburgh)
Theme: Data science challenges: tools from statistics to machine learning
Prompt detection of super-luminous supernovae with deep learning
Deep Learning is an area of Machine Learning that has gained great popularity in the last decade due to its success analyzing both images and sequential data. This makes it an ideal and powerful tool for time domain astronomy and specifically for the challenge of automatic transient detection and classification, that will be a relevant issue with the upcoming LSST data. Using simulated photometry, we attempt to perform light curve automatic classification as early as possible, without making use of any additional features (such as host galaxy photometric resdshift or handcrafted statistical features) other than the time series itself and place special importance in the correct detection of rare events such as Super Luminous Super Novae.
