A Brazilian student pursuing a Data Science Master's at ETH Zurich.
This website is designed to provide a clearer and more aesthetically pleasing view of my personal and professional journey.
Last updated on:
Hello, I'm Bruno Croso, a Brazilian student pursuing a Data Science Master's at ETH Zurich and currently working as a Research Assistant at the Professorship of Computational Social Science (COSS).
Before that, I graduated in Statistics from the University of São Paulo (USP). During my time at the Institute of Mathematics and Statistics, I dedicated myself to making the most of what it could provide me. With that in mind, as the top student in my class at the university, I actively pursued experiences in various domains, including junior enterprises, student weeks, and extension groups, often taking on leadership roles. Moreover, in order to explore realms beyond my institute, expand my academic horizons and contribute positively to society, I had the privilege of participating in academic research focused on improving Fake News detection models. Additionally, for the same reasons, I dedicated a year to volunteer as a mathematics teacher. Those experiences made me develop a deep passion for learning, as well as the desire to become a data scientist able to make a positive impact on the world.
Going beyond academic boundaries, I constantly seek new challenges and experiences. In my free time, I enjoy staying active through sports like canoeing, hiking and weight training. Furthermore, greatly influenced by an exchange carried out in high school, I have developed a profound love for traveling and immersing myself in the culture of different countries worldwide.
Research Assistant at the Professorship of Computational Social Science (COSS).
As a Data Science Intern at Farmtech, the leading digital rural credit company in Brazil, I primarily worked with SQL, Excel, and Power BI, as well as Python and R on a smaller scale, to develop comprehensive reports and studies to support commercial decision-making within the company.
As leader of the IME Jr Data Team, I had the opportunity to guide a group of 27 members in upskilling processes, diagnostic meetings, as well as in internal and external projects. Among my activities carried out, some were:
Participating as a Data Team Member at IME Jr, a student-run statistical, computational and educational consultancy, allowed me to dive into the world of practical data analysis. Among my activities at the company, I must mention:
I contributed as a volunteer Mathematics Teacher at Cursinho Quantum, a popular pre-college course, seeking to help democratize access to education by setting up and teaching weekly classes through the course of one year.
I collaborated in a team of four to develop a novel framework for analysing football tracking data at the ZSAC x SkillCorner Hackathon, organised by the Zurich Sports Analytics Club (ZSAC). Using SkillCorner tracking data from the men's friendly Switzerland vs. Germany, the project combined Graph Theory — specifically the Ford-Fulkerson Maximum Flow algorithm — with tracking-aware spatial metrics to quantify the continuous attacking threat of a possession sequence at any given moment of a match. By treating the attacking team as a directed capacity network, the model evaluates passing lanes, run corridors, and shot windows as probabilistic edges, producing a single threat score per frame that evolves dynamically throughout the game. The submission was awarded 2nd place by a jury comprising representatives from SkillCorner, UZH, the Swiss Football Association (SFV), and FIFA.
Having initially served as the President of the Organizing Committee for the University of São Paulo Statistics Week, I was humbled to continue contributing to this event as the Vice-President. In that role, my responsibilities evolved to focus on providing support, guidance, and strategic insights to the team. Collaborating closely with the committee, I leveraged my past experiences to assist in decision-making and offered advice to enhance the event's impact and reach.
As President of the Organizing Committee for the University of São Paulo Statistics Week, I had the privilege of leading a remarkable team of approximately 20 individuals in order to create an engaging event that brought students, faculty, and professionals together to explore more about the worlds of Statistics and Data Science. My role involved guiding the team in securing speakers and sponsors, as well as orchestrating logistics to ensure the event's success.
During my early years in university, I had the opportunity to learn more about the dynamic world of finance by participating in the IME-USP Financial Market League. During one semester, I had the privilege of attending exclusive classes and having insightful discussions in order to understand more about this field.
I was involved in an academic research under the guidance of Professor Roseli de Deus Lopes, investigating the incorporation of textual features into Graph Neural Networks (GNNs) for fake news detection on the social network X (formerly Twitter). Our research examined how integrating representations of user profile content and interactions involved in the spread of news into the graph's nodes influenced model performance. Additionally, we explored the impact of different embedders for generating these representations, comparing the widely-used static text encoder GloVe with the contextual encoder BERTweet. Finally, we also investigated whether upsampling combined with feature noise injection could enhance results, applying the NEFTune method with noise amplitudes of 0, 5, 10, 15, 20, and 25. All possible feature combinations were trained using the Graph Attention Network (GAT) architecture and the results were evaluated using the Wilcoxon signed-rank test for paired differences.
The extended abstract developed was accepted and presented as a poster at the Learning on Graphs Conference (LoG Conference) (2024) and is available at openreview.net.
For my second final project in my Bachelor's Degree, I collaborated as part of a duo responsible for the statistical analysis of the research project in question. This analysis sought to assess the influence of variables relating to the body composition of pregnant women in early pregnancy, such as the amount of fat mass in different parts of the body and BMI, on variables relating to fetal biometry and adiposity in the third trimester of pregnancy, such as fetal weight and cranial circumference. The study consisted of two main parts:
For my first final project in my Bachelor's Degree, I collaborated as part of a duo responsible for the statistical analysis of the research project in question. This analysis aimed to investigate changes in infant mortality in the State of Pernambuco by comparing the pre-pandemic and post-pandemic periods. The study consisted of three main parts:
Currently enrolled in the Master in Data Science programme.
Top ranked student in my entrance group with an average of 89%, recognised with the Academic Merit Diploma (CONRE-3 SP) and an Honourable Mention (IME-USP) as the best graduating student of the 2024-2 cohort.
Relevant modules: Introduction to Computer Science, Principles of Algorithm Design, Natural Language Processing, Computing and simulation laboratory, Differential and Integral Calculus I to IV, Linear Algebra I & II, Probability I & II, Statistical Inference, Multivariate and Nonparametric Inference, Multivariate Data Analysis, Regression Analysis, Bayesian Data Analysis, Time Series Analysis.
School where I did an exchange program to British Columbia, Canada. During this period, I had the opportunity to immerse myself in a new academic and cultural environment, sparking my interest in further international learning and experiences.
School where I completed the majority of my primary and secondary education, and where I first developed my passion for mathematics and statistics.
