Internship in Statistics & Data Science in the Digital Food Safety department. I worked on the development of a guideline for greenhouse gas (GHG) emissions estimation in the dairy sector. In particular, the objectives of my research were: 1. Derivation of an unbiased estimator for GHG emissions and the associated variance. 2. Development of a sampling design to optimize data collection. In order to accomplish such tasks, during the internship I focused on the following activities: - Review of scientific literature concerning GHG emissions estimation, Monte Carlo methods to compute variance, and advanced sampling designs. - Development of simulations using R programming language to assess the reliability of the methods employed in the guideline. - Writing a scientific report to gather all the information and results discovered during the research.

Working alongside neuroscientists at the City College of New York, providing statistical and analytical expertise in various Parkinson's disease and neuroscience research projects. In particular, I mostly worked on the following topics: PROJECT #1: Analysis of data related to patients affected by Parkinson's disease who followed the Art Therapy treatment Objective: To assess the effectiveness of the therapy to slow down the evolution of Parkinson's disease in patients at different stages. Data: General information and outcomes of tests assessing cognitive, motor, and visuospatial abilities of 50 patients, measured before and after the therapy (baseline and follow-up scores). Tasks: - Statistical analysis using nonparametric ANOVA tests to assess the effectiveness of the therapy - Development of a model to classify patients based on responsiveness using a pool of features selected via feature importance methods PROJECT #2: Co-authored in the paper: “Movement-related ERS and connectivity in the gamma frequency decrease with practice” Task: Developed specific non-parametric permutation tests and performed statistical analysis. PROJECT #3: Organizing hands-on Machine Learning workshops with Python for researchers in the Lab (the first one on EEG data analysis with Deep Learning will take place in Nov 2023).

Title: “Deep learning-driven image analysis of epithelial tissues structure and organization” Host: Computational Biology group at the D-BSSE @ ETH Zurich. The objective of the thesis project is to investigate the mechanism that drives the formation of different cell structures in epithelial tissues. I worked on the following topics: - Development of a deep learning-driven pipeline for the preprocessing and the segmentation of 3D microscope images of different epithelial tissues. - Implementation of an image analysis library to extract morphological statistics from segmented 3D cells. - Classification of cellular shape and morphology using a point cloud neural network called FoldingNet. - Implementation of a pipeline for 3D mesh generation and refinement. - Investigation of epithelial tissues' mechanical properties using a C++ 3D simulation framework based on cell meshes.