Astrophysics Source Code Library

Making codes discoverable since 1999

Photometric redshift estimation via deep learning

Discussion topics for individual codes
Ada Coda
ASCL Robot
Posts: 817
Joined: Thu May 08, 2014 5:37 am

DCMDN: Deep Convolutional Mixture Density Network

Postby Ada Coda » Wed Jul 12, 2017 2:47 pm

DCMDN: Deep Convolutional Mixture Density Network

Abstract: Deep Convolutional Mixture Density Network (DCMDN) estimates probabilistic photometric redshift directly from multi-band imaging data by combining a version of a deep convolutional network with a mixture density network. The estimates are expressed as Gaussian mixture models representing the probability density functions (PDFs) in the redshift space. In addition to the traditional scores, the continuous ranked probability score (CRPS) and the probability integral transform (PIT) are applied as performance criteria. DCMDN is able to predict redshift PDFs independently from the type of source, e.g. galaxies, quasars or stars and renders pre-classification of objects and feature extraction unnecessary; the method is extremely general and allows the solving of any kind of probabilistic regression problems based on imaging data, such as estimating metallicity or star formation rate in galaxies.

Credit: D'Isanto, Antonio; Polsterer, Kai Lars


Bibcode: 2017ascl.soft09006D

ID: ascl:1709.006
Last edited by Ada Coda on Thu Sep 07, 2017 9:43 pm, edited 1 time in total.
Reason: Updated code entry.

Return to “Astrophysics Source Code Library”

Who is online

Users browsing this forum: Bing [Bot] and 0 guests