Emanuele Iaccarino personal mark Emanuele Iaccarino personal mark E. Iaccarino
  • home
  • about
  • academic
  • projects
  • publications
  • awards & certifications
  • CV
  • get in touch!
    • LinkedIn
    • email

On this page

  • Overview
  • Main objectives
  • Learning outcomes
  • My work in this course

Statistical Consulting

University of Naples Federico II · B.Sc. in Statistics for Business and Society

Published

March 1, 2024

Overview

This was a practice-oriented course designed to expose students to real-world problems in Machine Learning and Artificial Intelligence.
Unlike other highly theoretical courses, it emphasized hands-on experimentation, creative problem-solving, and direct engagement with algorithms applied to real datasets.

The professor gave students space to explore and often encouraged deeper dives into more advanced topics, which strongly influenced my academic trajectory and motivated my decision to pursue further studies in Rome.


Main objectives

  • Learn to apply machine learning algorithms in practice rather than only studying their theory
  • Gain first-hand experience with clustering methods and supervised learning techniques
  • Explore boosting algorithms, hyperparameter tuning, and ensemble methods
  • Develop critical problem-solving skills for tackling imbalanced data and real-world complexity

Learning outcomes

  • Ability to implement and tune supervised and unsupervised algorithms
  • Understand the role of boosting and ensembling in improving performance
  • Apply modern tools for model optimization (e.g., Optuna for hyperparameter tuning)
  • Experiment with data augmentation techniques, including GANs for synthetic data generation
  • Translate Kaggle-like competition setups into structured learning experiences

My work in this course

  • Exploratory projects
    • RAG Chatbot – Local PDF QA
    • Animal Emotion Recognition via LiveCam
  • Final Project
    • Analysis on Churn Banking Dataset
    • Ensemble of boosting algorithms with hyperparameter tuning (Optuna)
    • Cost-sensitive learning to handle data imbalance
    • Synthetic data generation with GANs

© 2025 Emanuele Iaccarino

 

This website is built with , , and Quarto