Bringing Manifold Learning and Dimensionality Reduction to SED Fitters

Research output: Contribution to journal › Journal article › Research › peer-review

Standard

Bringing Manifold Learning and Dimensionality Reduction to SED Fitters. / Hemmati, Shoubaneh; Capak, Peter; Pourrahmani, Milad; Nayyeri, Hooshang; Stern, Daniel; Mobasher, Bahram; Darvish, Behnam; Davidzon, Iary; Ilbert, Olivier; Masters, Daniel; Shahidi, Abtin.

In: Astrophysics Journal Letters, Vol. 881, No. 1, L14, 01.08.2019.

Research output: Contribution to journal › Journal article › Research › peer-review

Harvard

Hemmati, S, Capak, P, Pourrahmani, M, Nayyeri, H, Stern, D, Mobasher, B, Darvish, B, Davidzon, I, Ilbert, O, Masters, D & Shahidi, A 2019, 'Bringing Manifold Learning and Dimensionality Reduction to SED Fitters', Astrophysics Journal Letters, vol. 881, no. 1, L14. https://doi.org/10.3847/2041-8213/ab3418

APA

Hemmati, S., Capak, P., Pourrahmani, M., Nayyeri, H., Stern, D., Mobasher, B., Darvish, B., Davidzon, I., Ilbert, O., Masters, D., & Shahidi, A. (2019). Bringing Manifold Learning and Dimensionality Reduction to SED Fitters. Astrophysics Journal Letters, 881(1), [L14]. https://doi.org/10.3847/2041-8213/ab3418

Vancouver

Hemmati S, Capak P, Pourrahmani M, Nayyeri H, Stern D, Mobasher B et al. Bringing Manifold Learning and Dimensionality Reduction to SED Fitters. Astrophysics Journal Letters. 2019 Aug 1;881(1). L14. https://doi.org/10.3847/2041-8213/ab3418

Author

Hemmati, Shoubaneh ; Capak, Peter ; Pourrahmani, Milad ; Nayyeri, Hooshang ; Stern, Daniel ; Mobasher, Bahram ; Darvish, Behnam ; Davidzon, Iary ; Ilbert, Olivier ; Masters, Daniel ; Shahidi, Abtin. / Bringing Manifold Learning and Dimensionality Reduction to SED Fitters. In: Astrophysics Journal Letters. 2019 ; Vol. 881, No. 1.

Bibtex

@article{1e4e42a33525419d94e19f3fc9b63c4b,
title = "Bringing Manifold Learning and Dimensionality Reduction to SED Fitters",
abstract = "We show that unsupervised machine learning techniques are a valuable tool for both visualizing and computationally accelerating the estimation of galaxy physical properties from photometric data. As a proof of concept, we use self-organizing maps (SOMs) to visualize a spectral energy distribution (SED) model library in the observed photometry space. The resulting visual maps allow for a better understanding of how the observed data maps to physical properties and allows for better optimization of the model libraries for a given set of observational data. Next, the SOMs are used to estimate the physical parameters of 14,000 z ˜ 1 galaxies in the COSMOS field and are found to be in agreement with those measured with SED fitting. However, the SOM method is able to estimate the full probability distribution functions for each galaxy up to ˜106 times faster than direct model fitting. We conclude by discussing how this acceleration, as well as learning how the galaxy data manifold maps to physical parameter space and visualizing this mapping in lower dimensions, helps overcome other challenges in galaxy formation and evolution.",
keywords = "galaxies: fundamental parameters, galaxies: statistics",
author = "Shoubaneh Hemmati and Peter Capak and Milad Pourrahmani and Hooshang Nayyeri and Daniel Stern and Bahram Mobasher and Behnam Darvish and Iary Davidzon and Olivier Ilbert and Daniel Masters and Abtin Shahidi",
year = "2019",
month = aug,
day = "1",
doi = "10.3847/2041-8213/ab3418",
language = "English",
volume = "881",
journal = "The Astrophysical Journal Letters",
issn = "2041-8205",
publisher = "IOP Publishing",
number = "1",

}

RIS

TY - JOUR

T1 - Bringing Manifold Learning and Dimensionality Reduction to SED Fitters

AU - Hemmati, Shoubaneh

AU - Capak, Peter

AU - Pourrahmani, Milad

AU - Nayyeri, Hooshang

AU - Stern, Daniel

AU - Mobasher, Bahram

AU - Darvish, Behnam

AU - Davidzon, Iary

AU - Ilbert, Olivier

AU - Masters, Daniel

AU - Shahidi, Abtin

PY - 2019/8/1

Y1 - 2019/8/1

N2 - We show that unsupervised machine learning techniques are a valuable tool for both visualizing and computationally accelerating the estimation of galaxy physical properties from photometric data. As a proof of concept, we use self-organizing maps (SOMs) to visualize a spectral energy distribution (SED) model library in the observed photometry space. The resulting visual maps allow for a better understanding of how the observed data maps to physical properties and allows for better optimization of the model libraries for a given set of observational data. Next, the SOMs are used to estimate the physical parameters of 14,000 z ˜ 1 galaxies in the COSMOS field and are found to be in agreement with those measured with SED fitting. However, the SOM method is able to estimate the full probability distribution functions for each galaxy up to ˜106 times faster than direct model fitting. We conclude by discussing how this acceleration, as well as learning how the galaxy data manifold maps to physical parameter space and visualizing this mapping in lower dimensions, helps overcome other challenges in galaxy formation and evolution.

AB - We show that unsupervised machine learning techniques are a valuable tool for both visualizing and computationally accelerating the estimation of galaxy physical properties from photometric data. As a proof of concept, we use self-organizing maps (SOMs) to visualize a spectral energy distribution (SED) model library in the observed photometry space. The resulting visual maps allow for a better understanding of how the observed data maps to physical properties and allows for better optimization of the model libraries for a given set of observational data. Next, the SOMs are used to estimate the physical parameters of 14,000 z ˜ 1 galaxies in the COSMOS field and are found to be in agreement with those measured with SED fitting. However, the SOM method is able to estimate the full probability distribution functions for each galaxy up to ˜106 times faster than direct model fitting. We conclude by discussing how this acceleration, as well as learning how the galaxy data manifold maps to physical parameter space and visualizing this mapping in lower dimensions, helps overcome other challenges in galaxy formation and evolution.

KW - galaxies: fundamental parameters

KW - galaxies: statistics

U2 - 10.3847/2041-8213/ab3418

DO - 10.3847/2041-8213/ab3418

M3 - Journal article

VL - 881

JO - The Astrophysical Journal Letters

JF - The Astrophysical Journal Letters

SN - 2041-8205

IS - 1

M1 - L14

ER -

ID: 236165052