Big Data analysis techniques applied to simulated data for the preparation of the space mission PLATO
SECOND CALL
APPLICATION ID: ALL11
What we are looking for:
We are seeking innovative research proposals that explore Artificial Neural Networks (ANNs) as a pioneering and breakthrough way to interpret observations by combining theoretical stellar models, stellar population models, and prescriptions for observational biases. The selected PostDoc will join a multidisciplinary research group between the Instituto de Astrofísica de Andalucía y el Instituto de Ciencias de Espacio.
The context:
PLATO is a M3 space mission from the European Space Agency which is aimed to detect and characterize Earth analogs and their hosting stars. The huge amount of data that will be gathered is on a par with the stellar models and theoretical simulations required to extract the information. The development of Artificial intelligence (AI) systems provide a unique opportunity to exploit both observations and models and push the current limits of our knowledge of stars and planets.
The problem to address:
Javier Pascual, J.R. Rodón (IAA), and Aldo Serenelli (ICE) are joining efforts to implement a promising holistic approach where AI is used to extract the physical information both from observations and stellar models in preparation for the PLATO space mission, which will be launched at the end of 2026. The proposal will be developed in collaboration with 2 partners which are research groups of international excellence. Universities: Centro de Astrofísica da Universidade do Porto (CAUP, PT); LESIA, Paris Observatory (FR).
Objectives:
- To develop a prototype of ANN that can detect anomalies in datasets based on prescriptions for avoiding observational biases, and demonstrate the practical application in various scenarios..
- To enhance state-of-the-art forward modelling in stellar astrophysics through the implementation of ANNs as a bridge between the unbiased analysis of observations and theoretical models of stellar evolution and pulsation.
Qualifications:
-
Indespensable: PhD preferably in Physics or AI
-
Solid knowledge of Artificial Intelligence techniques
-
Some expertise in modelling physical processes
-
High level computer programming skills, e.g. Python
-
Advanced statistical knowledge
-
Other desirable skills: data analysis, analytical thinking, problem solving, communication skills
Expected Outcomes:
- Build and train an ANN aligning observational and theoretical representations of stellar physical processes.
- Applications of the ANN both to improve data analysis of stellar pulsations and current limitations of theoretical models.
