Federico Marulli

Available thesis projects for the First Cycle Degrees in Astronomy or Physics

1) Comparing alternative estimators of the two-point and three-point correlation functions

Goal: implement and validate alternative statistical estimators of second-order and third-order clustering statistics

Prospects:

• learn how to use standard codes for cosmological analyses
• contribute to the CosmoBolognaLib implementation
• acquire numerical knowledges, in particular in C++ and Python, useful for both scientific and many other non-scientific activities

• scientific skills: low
• computing skills: medium (C++/Python)

2) A graphic web interface for the CosmoBolognaLib

Goal: expand the web interface of the CosmoBolognaLib to include further cosmological probe statistics

Prospects:

• learn how to implement graphic web interfaces
• build new cosmological tools for general usage
• familiarize with Python, one of the most popular programming languages

• scientific skills: low
• computing skills: medium (Python)

3) Emulating cosmological observable

Goal: train deep neaural networks to emulate cosmological probe observables

Prospects:

• become expert in deep learning techniques
• acquire high-level cosmological knowledges, useful for the Master Thesis
• contribute to the optimization of the CosmoBolognaLib
• acquire high-level skills in C++ and Python, useful for both scientific and many other non-scientific activities
  

• scientific skills: medium
• computing skills: medium (C++/Python)

Available thesis projects for the Second Cycle Degree in Astrophysics and Cosmology

1) Validating the Euclid clustering codes on the Flagship Simulation

Goal: test the new Euclid clustering pipeline on galaxy and cluster mock Euclid catalogues

Prospects:

• contribute to the preparation of the ESA Euclid mission
• become a C++ expert
• learn how to exploit the latest numerical tools to do cosmology with next generation galaxy surveys
• contribute to write internal Euclid reports

Requirements:

• scientific skills: medium
• computing skills: medium/high (C++)
  

2) Cosmological constraints from cluster number counts and clustering in Euclid

Goal: participate to the Euclid Cluster Cosmology Challenge, aimed at implementing the Euclid likelihood modules to model cluster statistics

Prospects:

• contribute to the preparation of the ESA Euclid mission
• become a C++/Python expert
• learn how to exploit the latest numerical tools to do cosmology with next generation cluster surveys
• contribute to implement new Euclid modules for cosmological analyses
• contribute to write internal Euclid reports

Requirements:

• scientific skills: medium
• computing skills: high (C++/Python)

3) Halo Occupation Distribution model to extract cosmological constraints from small-scale clustering

Goal: implement the Halo Occupation Distribution model in the CosmoBolognaLib, and exploit it on real data

Prospects:

• become one of the CosmoBolognaLib main builders
• become a C++/Python expert
• derive new cosmological constraints from galaxy clustering at small scales (possible applications to VIPERS and BOSS surveys)
• contribute to write one or more scientific publications

Requirements:

• scientific skills: medium
• computing skills: high (C++/Python)

4) Combining high and low redshift cosmological probes

Goal: collect a large dataset of cosmological data from different probes, and implement the numerical tools to combine them

Prospects:

• acquire high-level cosmological knowledges on different probes
• acquire high-level statistical skills
• become a C++/Python expert
• familiarize with the latest data analysis techniques, useful for both scientific and many other non-scientific activities
• contribute to write one or more scientific publications

Requirements:

• scientific skills: high
• computing skills: medium/high (C++/Python)

5) Bayesian deep neural networks to learn the properties of the Cosmic Web

Goal: exploit the latest machine learning techniques to do cosmology bypassing standard statistics

Prospects:

• become an expert on deep learning techniques
• develop new methods, never tested before, for cosmological analyses
• become one of the first builders of the CosmoBolognaNet
• start a long-term project, to be hopefully continued during the Ph.D.

Requirements:

• scientific skills: high
• computing skills: high (Python)