Data Science

Risultati di apprendimento attesi: Graphs have applications in multiple areas, including social networks, bioinformatics, network medicine, computational chemistry, and they can be used to provide tools in these areas. The course will present models and algorithms for the analysis of graphs as with applications on various areas. The goal at the end of the course, is for student to know algorithms and frameworks that can allow them to analyze large graph data. Informazioni sui prerequisiti culturali e curriculari necessari - Knowledge of basic algorithms - Programming - Linear algebra - Probability - Neural networks Programma in italiano • Theoretical algorithms for graph modeling and analysis: ◦ Real graph properties and models (Gnp, preferential attachment, Kleinberg’s reachability) ◦ Models for propagation (linear threshold, cascade) and for opinion formation ◦ Homophily and influence and algorithms for identifying and distinguishing ◦ Influence maximization ◦ Algorithms for graph alignment ◦ Dense subgraphs, community detection, graph minors ◦ Graph summarization and sampling • Machine-learning approaches: ◦ Label propagation ◦ Graph transformers ◦ Knowledge-graph emdeddings ◦ Models for analysis of temporal graphs ◦ Explainability • Architectures for handling large graph data: ◦ Spark GraphsX ◦ AWS Neptune ◦ AWS GraphStorm ◦ Neo4J Modalità di valutazione delle conoscenze Prova scritta Prova orale Valutazione progetto Modalità di valutazione in italiano Homeworks and/or project and oral exam or written exam Esempi di domande e/o esercizi frequenti Find the most influential nodes in a network. Testi adottati Material will be distributed online Modalità di svolgimento Didattica frontale/tradizionale Modalità di svolgimento in italiano The course is based on in-class theoretical lectures and sometimes in-class labs. Modalità di frequenza Classes are in person. Programmazione: http://aris.me/index.php/teaching