From Science to Data Science
I am a theoretical physicist who worked more and more with data, using them to test theories. I now typically work in roles at the interface between theory and data. Since 2016 I got interested in machine learning.
Here you can find more details on specific data science projects I worked on and relevant training certificates.
2016 - 2020
Python programming language
The EuroPython Society (EPS) is a Swedish non-profit organization which supports the Python community in Europe. The annual EuroPython Conference involves about 1200 attendees, from academia and industry, interested in a variety of applications of python programming language. I was invited speaker in 2016 and liked the community so much that I volunteered to help with the organisation, since 2017. In 2018-2019 I was elected in the EuroPython Society Board. In 2019-2020 I helped with EuroSciPy conference.
Distinguishing standard and modified gravity cosmologies with machine learning
Convolutional Neural Networks in cosmology
In this work (Peel et al 2019) we used a CNN to classify distinct cosmological scenarios based on different weak lensing maps they generate. We correctly identify among the modified gravity models with at least 80% accuracy when using the full redshift information.
Science to Data Science
Healthcare Internet of Things
I am Alumna of S2DS, Europe’s largest data science training programme. Run by Pivigo, it consists in five weeks of intensive, project-based training turning Scientists into Data Scientists. During my project-based training, I worked in a team for an Internet of Things (IOT) project on healthcare. Given different databases (simulated and real, in different formats) on patients, labs and health parameters (pressure, blood, ...) I had to identify rules and correlation among parameters to alert the patient of the potential onset of a specific disease. The main software language used for the project was python.
Planck 2015 results. XIV. Dark energy and modified gravity
Monte Carlo simulations
I have 10+ years experience in using Monte Carlo simulations and statistics to compare theory predictions with data coming from galaxy surveys, and from the cosmic microwave background. Therabytes of data are compressed, preprocessed, masked, to extract cosmological parameters that best describe the content and evolution of the universe.
S2DS Virtual March 2016
"Europe’s largest data science training programme. Five weeks of intensive, project-based training turning exceptional analytical PhDs and MScs into Data Scientists. S2DS is designed to help all participants to use their analytical experience and apply it to a real world problem with one of our partner companies" More info: http://www.s2ds.org/ During my training I worked on applying machine learning to a healthcare IOT problem, for a startup based in London.