PhD student in astronomical classification with 4MOST

Submission Information
Publish Date: 
Tuesday, October 4, 2022
Archive Date: 
Tuesday, November 1, 2022
Job Summary
Job Category: 
Pre-doctoral/Graduate Positions
Institution Classification/Type: 
Foreign
Institution/Company: 
Uppsala University
Department Name: 
Department of Physics and Astronomy
City: 
Uppsala
State/Province: 
Uppsala
Country: 
Sweden
Announcement
Job Announcement Text: 

<p>The Division for Astronomy and Space Physics at Uppsala University is hiring a PhD student in astronomy, focused on object classification of astronomical sources via machine learning with the telescope 4MOST (4-metre Multi-Object Spectroscopic Telescope).</p>

<p><strong>Duties&nbsp;</strong><br />
    The PhD student is expected to work on the development and implementation of astrostatistical machine learning methods for automatic classification and anomaly detection in analyses of complex and large astronomical data sets. The work will be performed in the context of the scientific collaboration around the telescope 4MOST (4-metre Multi-Object Spectroscopic Telescope,&nbsp;http://www.4most.eu).</p>

<p>4MOST will be Europe&#39;s new workhorse for astronomical spectroscopy. Starting in 2024, 4MOST will collect optical spectra of 40 million objects, both Galactic and extra-galactic ones. Uppsala University has been a science partner in the project from the start. To maximize the science output and the potential for discovery, we require novel and efficient methods for object classification and anomaly detection, that do not rely on pre-existing theoretical or empirical models. The full 4MOST data set will contain around three trillion data points. This necessitates a probabilistic classification method which can be repeatedly and efficiently re-trained.</p>

<p>The goal of the PhD project is to develop an efficient automatic and dynamically trainable classification and anomaly detection pipeline for spectroscopic + photometric astronomical surveys, which can compute probabilities for each object (astronomical source) in very large data sets to belong to different categories of objects (e.g. stars, galaxies, active galactic nuclei). To achieve this, supervised and unsupervised Bayesian machine learning methods will be adapted and employed, as well as methods for handling vast and high-velocity data sets. Initially, the work will be performed using data from earlier astronomical surveys and simulations, and is expected to transition to using observational data from 4MOST in 2024/25.</p>

<p>The PhD student will belong to the eSSENCE &ndash; SciLifeLab graduate school in data-intensive science, and is expected to participate and collaborate within the graduate school.</p>

Application Deadline: 
Tuesday, November 1, 2022
Current Status of Position: 
Accepting Applicants
Apply to Job
Attention To: 
Martin Sahlén
Institution/Company Job ID or Reference Code: 
UFV-PA 2022/3553
Inquiries About Job
Attention To: 
Martin Sahlén
Subject: 
Query re UFV-PA 2022/1263