We have a unique opportunity to make a definitive step forward and conclusively reconstruct the life of a typical (super) Milky-Way galaxy, from its initial clumpy gas-rich phase to its (inevitable?) death through quenching. Astronomy is indeed entering the big data era. Big-data spectro-photometric surveys such as LSST, EUCLID, WFIRST, JWST will be on sky in 2-5 years from now. While most emphasis is being put on the cosmology aspects, invaluable information on the star-formation and assembly histories of billions of galaxies across 80% of the cosmic time will be encoded in these datasets. Extracting and interpreting this information is a new challenge for which no valid solution has been provided yet. ASTROBRAIN is thus motivated by the evidence that new techniques need to be invented now to fully benefit from these datasets and advance in our knowledge of galaxy formation.
I therefore ask support to start building a new inter-disciplinary team at the IAC. - ASTROBRAIN - that will set the route for the new era in astrophysics science by:

-- Developing a novel analysis of spectro-photometric data through with a core of artificial intelligence. The proposed methodology will go beyond state-of-the-art techniques by allowing, for the first time, to reconstruct the mass assembly histories of a massive sample of galaxies (WP3 - e.g. interactions, bulge growth, feedback, instabilities etc) which will be then linked to star-formation and dark-matter. We will that way provide a comprehensive view of how massive galaxies evolve over 80% of the cosmic time. In order to achieve this goal, we will build a unique database of outputs of state-of-the art hydrodynamic numerical simulations (e.g. IlustrisTNG, HorizonAGN, VELA) projected in the observational space (WP1). It will constitute a unique tool for comparisons of observations and simulations and also between simulations.


-- Ensuring an invaluable legacy value. It will make the definitive and unavoidable step of replacing humans by computers in the analysis of astronomical data and will prepare the community, specially young researchers, for the new big-data era in astronomy. We will develop an end-to-end AI based pipeline to process imaging data from deep surveys (WP2). All developed tools to extract assembly histories from big- data surveys will be also released as a legacy product. Expected to become a new standard for extragalactic imaging surveys, these tools will be of tremendous impact for other groups in the institute and for our community for all planed future scheduled extra-large surveys such as LSST, EUCLID, WFIRST and also JWST.