Cambridge, England, United Kingdom
I am a Ph.D. Candidate at the Institute of Astronomy, University of Cambridge, fully funded by a "la Caixa" Postgraduate scholarship and a Centre for Doctoral Training in Data Intensive Science scholarship through Wolfson College. My Ph.D. research is supervised by Prof. Vasily Belokurov and Prof. Miles Cranmer. I graduated with a double BSc in Physics and a BSc in Mathematics from the Autonomous University of Barcelona, fully sponsored by a Pere Menal Scholarship. My research interests include astronomy, cosmology, and the application of machine learning and data science to these fields. Feel free to reach out if you’d like to discuss any work-related topics!
Supervisor for the following courses: M1: Machine Learning M2: Deep Learning
· Conducted a performance assessment of the upcoming Cherenkov Telescope Array Observatory (CTAO). · Established estimated expected performance improvements at different construction stages. · Devised a python pipeline to study the estimated best sub-array candidate for monitoring different AGNs and variable sources. · Based on Monte Carlo simulations, Instrument Response Functions computations and simulations of variable sources spectras and detections. · Attended advanced lectures in particle physics in parallel with my work.
· Grade: 100% · Implementation and hyperparameter tuning of the TabNet (encoder architecture) algorithm to fit tabular data classification models for commercial campaings. · Compared performance results between TabNet and XGBoost (Extreme Gradient Boosting) for high-monetary-gain data sets. · Obtained better performance results using the TabNet algorithm (Neural Network), with hyperparameter tuning, for the larger sized training data sets (above 200 MB).
· Curricular (3 months) and extracurricular (3 months). · Data scientist intern in the department Advanced Finantial Modelling, Artificial Intelligence and Machine Learning. · Established the training perimeter, test and valida data sets for a solar panel purchase classification model. Commercial implementation. · Automatised the creation of large-scale tables from SAS to SQL and Python.
· Grade: 95%, A+ · Analysed the statistics of close encounters between nearby stars using Gaia data. Exoplanet detection via the Radial Velocity technique and stellar activity indices using CARMENES data.