Projects

SOME OF MY WORKS

Development and evaluation of automated waveform analysis techniques for BEGe detectors within the VIP project. Two approaches were tested: feature-based machine learning and deep learning directly on raw signals. A custom CDAE–FAE–GMVAE neural architecture achieved up to 98% accuracy, enabling efficient pulse discrimination even with minimal labeled data. Download the Slides

A data-driven investigation of how climate change affects regional economic and social inequalities between northern and southern Italy, using mixed-effects models, data visualization, and climate-economy predictors. Download the full Report

A comparative study of PCA and GPLVM applied to a simulated cylindrical dataset and the MNIST handwritten digits, evaluating the performance of both dimensionality reduction techniques. This project was developed as part of my bachelor’s thesis.
PDF

Comparative analysis of Italian legislation on abortion and conscientious objection, highlighting similarities and differences with other European countries. Includes data visualization of abortion and conscientious objection rates and a customized map created in RStudio.
PDF

Investigation of how house prices in Portsmouth are influenced by property characteristics, analyzed through statistical modelling, a Shiny App, and a poster presentation. Download the Poster or explore the Shiny App

Conferences and Talks

MY CONTRIBUTION

Apr 9, 2024 Wellesley College, Boston, Massachusetts (USA)

I was part of the Ruhlman Conference, held on 11 April 2024 at Wellesley College in Massachusetts, together with Professor Daniela Cavalcoli, delegated by the Rector of the Alma Mater Studiorum - Università di Bologna for the mobility of lecturers and students. One of the Ruhlman Conference’s goals is promoting research through interdisciplinary collaboration of students and lecturers, to encourage the enrichment of the College intellectual life. The Conference had an international outlook. In fact, six other students and I were selected from universities all over the world. During the conference I had the opportunity to share the results and experiences of my research, which investigates how the use of two different dimensionality reduction techniques (PCA and GPLVM), applied to a simulated dataset and a real dataset of handwritten digits, impacts on the quality of the latent spaces obtained. Attending the international conference provided new inspiration for the continuation of my journey. A special thanks to the Wellesley College staff for organizing the Conference and hosting us, and to Professor Daniela Cavalcoli for sharing this wonderful experience with me.

Soft kills

TRAITS THAT SET ME APART

Competence

Leadership

Reliability

Focus

Education

 
 
 
 
 

Master in Computational Sciences (Double Degree)

Università della Svizzera Italiana (USI)

Sep 2025 – Present Lugano, Switzerland
The Master in Computational Sciences at USI provides advanced interdisciplinary training in statistical modelling and computational methods.
Coursework includes Machine Learning, Deep Learning, Bioinformatics, and Network Analysis.
 
 
 
 
 

Master in Statistical Sciences (Curriculum: Data Science)

Alma Mater Studiorum – Università di Bologna

Sep 2024 – Present Bologna, Italy
The Master in Statistical Sciences focuses on advanced statistical modelling, computational inference, and applied data science.
 
 
 
 
 

Bachelor of Science in Statistics (Double Degree)

University of Glasgow

Sep 2023 – Jun 2024 Glasgow, United Kingdom
Completed within the Double Degree Programme with the University of Bologna.
Focused on statistical modelling, bayesian methods, mixed effect models, functional data analysis, and data visualization.
Gained hands-on experience with RStudio, ggplot2, and tidyverse, emphasizing applied data interpretation and communication.
 
 
 
 
 

Bachelor in Statistical Sciences (Stats & Maths)

Alma Mater Studiorum – Università di Bologna

Sep 2021 – Jul 2024 Bologna, Italy
A highly quantitative degree combining mathematics, statistics, and data analysis.
Final grade: 110/110 cum Laude
Thesis: Dimension Reduction: PCA and GPLVM for Simulated Cylinder and Real Handwritten Digits
 
 
 
 
 

Collegio Superiore – Honours Programme

Alma Mater Studiorum – Università di Bologna

Sep 2021 – Jul 2025 Bologna, Italy

Collegio Superiore is the honours programme of the University of Bologna, dedicated to a small number of highly selected students (≈5% acceptance rate, through written and oral exams).
It provides interdisciplinary courses, seminars, and mentorship opportunities that complement the main degree programme.

The programme also grants housing and a full-ride scholarship, conditional on maintaining a GPA of at least 28/30 in both the main degree and Collegio Superiore courses.

Work Experience

 
 
 
 
 

Teaching Assistant for Introduction to Data Science (MSc)

Università della Svizzera Italiana (USI)

Sep 2025 – Present Lugano, Switzerland

Responsibilities include:

  • Supervising master-level Data Science tutoring sessions of 10+ students.
  • Assisting students with statistical exercises and data analysis in Python.
  • Providing clear explanations of theoretical and computational methods.
  • Monitoring student progress and offering individualized support.
  • Communicating feedback to course coordinators about student performance.
    Competences: Data Science · Statistics · Teaching Skills · Python · Communication Skills
 
 
 
 
 

Summer Internship – VIP Experiment: Pulse Shape Analysis

INFN – Istituto Nazionale di Fisica Nucleare

Jul 2025 – Sep 2025 Frascati (Rome), Italy

Project: VIP Experiment – Automatic pulse shape analysis with BEGe detectors.
Main activities:

  • Developed algorithms for waveform feature extraction and classification.
  • Applied classical ML models (Random Forest, Gradient Boosting) with strong benchmark results.
  • Designed a deep learning pipeline (CDAE+FAE+GMVAE) for semi-supervised classification using TensorFlow/Keras.
  • Achieved up to 98% accuracy with limited labeled data, demonstrating model robustness.
  • Enhanced research efficiency by reducing manual labeling requirements.
    Competences: Signal Processing · Machine Learning · Deep Learning · Python · Data Analysis · Experimental Physics
 
 
 
 
 

Tutor for Time Series in Statistics

Alma Mater Studiorum – University of Bologna

Feb 2025 – Present Bologna, Italy

Responsibilities include:

  • Supervising undergraduate tutoring sessions for the Time Series course.
  • Guiding students through applied exercises in RStudio.
  • Clarifying statistical modelling concepts and problem-solving strategies.
  • Monitoring learning progress and providing targeted academic support.
    Competences: Time Series · Data Science · RStudio · Teaching Skills · Communication Skills
 
 
 
 
 

Mathematics Tutor

International School of Bologna

Sep 2024 – Jul 2025 Bologna, Italy

Responsibilities include:

  • Providing focused tutoring to Grade 12 students in calculus, statistics, algebra, and trigonometry.
  • Preparing students for final and university entrance exams (SAT, IB Mathematics).
  • Customizing lesson plans to individual needs and progress.
  • Supporting critical thinking and problem-solving development.
  • Collaborating with teachers to evaluate academic growth.
    Competences: Teaching Skills · Mathematics · Communication · Educational Support
 
 
 
 
 

Tutor & Lab Demonstrator for Statistics

University of Glasgow

Oct 2023 – Jun 2024 Glasgow, United Kingdom

Roles: Tutor and Lab Demonstrator for undergraduate Statistics courses.
Responsibilities:

  • Conducted weekly tutorials and lab sessions (10–30 students per group).
  • Supported students using RStudio for applied data analysis.
  • Clarified theoretical concepts and statistical methods.
  • Provided feedback to coordinators on class progress and learning outcomes.
    Competences: RStudio · ggplot2 · Tidyverse · Data Science · Algorithms · Teaching · Communication Skills

Contact

Connect with me

  • ivan.carnevali02@gmail.com
  • +39 3207759861
  • Piazza Vittorio Puintoni 1B