Foundations of Ai

The cluster aims to provide students with foundational conceptual and practical knowledge in artificial intelligence and data analysis, with the goal of understanding how data can be collected, prepared, explored, and used to build basic intelligent systems.
The educational pathway integrates data analysis and visualization, programming, and the foundations of machine learning, guiding students from understanding datasets and their critical issues to building and evaluating models, and finally to reading and communicating results. The cluster develops essential skills to properly frame a data-driven problem, select appropriate techniques, and implement reproducible pipelines.
The cluster is organized into the modules of Data Analysis and Visualization, Programming and Algorithms II, and Machine Learning, designed in a coordinated manner. The modules jointly contribute to the intended learning outcomes following a pipeline logic: data exploration and representation, implementation of reproducible procedures and experiments, and training and evaluation of supervised and unsupervised models.

Knowledge and understanding
At the end of the cluster, the student acquires foundational knowledge of data, descriptive statistics, visualization, and machine learning, which is necessary to understand the main phases of a data-driven process. In particular, the student is able to:
· describe the key stages of a data-driven pipeline (collection, cleaning, exploration, modelling, evaluation, and communication);
· recognize common data types and basic descriptive statistics concepts used for initial exploration;
· describe fundamental concepts of supervised and unsupervised learning and the main classes of introductory models;
· explain the role of evaluation metrics and validation principles, understanding the importance of a sound assessment of results.
Applying knowledge and understanding
At the end of the cluster, the student is able to:
· perform descriptive and exploratory analyses, identifying patterns, outliers, and dataset issues and representing them through appropriate visualizations;
· implement, using programming tools and standard libraries, basic data preparation procedures and reproducible training/evaluation pipelines;
· train and compare basic machine learning models (supervised and unsupervised) on assigned datasets, selecting metrics appropriate to the task;
· apply introductory validation procedures (e.g., train/test split and cross-validation in guided settings) to reduce biased evaluations.
Making judgements
The student develops the ability to:
· select basic analysis techniques and models consistent with the objective and data characteristics;
· critically interpret results (strengths, limitations, errors), recognizing common signs of unreliable evaluation (e.g., overfitting or inappropriate metric choices);
· identify potential improvement strategies (pre-processing, feature selection, model and parameter choices) and justify their expected impact.
Communication skills
At the end of the cluster, the student is able to:
· communicate clearly and in a technically correct manner the addressed problem, the adopted pipeline, the obtained results, and main uncertainties/limitations, using concise reports and explanatory visualizations.
Learning skills
The student acquires the ability to:
· independently consult documentation and technical resources (libraries, tools, methodological guidelines) to extend or adapt analysis and modelling pipelines;
· replicate and adapt experiments on new datasets by defining coherent evaluation criteria and checking the robustness of conclusions;
· learn and use new basic models or tools by leveraging previously acquired mathematical and programming skills;
· plan subsequent iterations of the analysis (improvements in data, models, and validation) based on the observed results.