From e3c156eb1e3d4e1129334b0939e948b2eb54015a Mon Sep 17 00:00:00 2001 From: Andrei Igoshev Date: Fri, 4 Sep 2020 15:52:22 +0100 Subject: [PATCH] Machine learning applied to open clusters --- astrohackweek20/random-forest-open-cluster.yml | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) create mode 100644 astrohackweek20/random-forest-open-cluster.yml diff --git a/astrohackweek20/random-forest-open-cluster.yml b/astrohackweek20/random-forest-open-cluster.yml new file mode 100644 index 0000000..83b9f64 --- /dev/null +++ b/astrohackweek20/random-forest-open-cluster.yml @@ -0,0 +1,16 @@ +title: Machine learning applied to open cluster identification +creators: + - Andrei Igoshev + - Catarina Alves + - Ernest Glukhov + - Marcin Marculewicz + - Mariia Demianenko + - Tarun Ruchandani +description: > Code to verify if a collection of stars found in the Gaia database is an open cluster. +Clustering algorithms (e.g. DBSCAN) returns tens of cluster candidates for every few square degrees +of the sky area when they are applied to the Gaia DR2 database data. In reality some of these cluster +candidates are not open clusters but rather are fluctuations of the Galactic disk stellar population. +These collections of stars are expected to have very different ages, absorptions and metallicities. +In this repository we train random forest and convolutional neural network to distiungush between +actual open cluster and general Galactic disk population. +source-url: https://github.com/ignotur/Random-forest-open-cluster