Etibar Vazirov

Dottorando in Ingegneria Informatica E Dei Sistemi , 40o ciclo (2024-2027)
Dipartimento di Automatica e Informatica (DAUIN)

Docente esterno e/o collaboratore didattico
Dipartimento di Automatica e Informatica (DAUIN)

Profilo

Dottorato di ricerca

Argomento di ricerca

Graph Neural Networks for Data Science

Tutori

Interessi di ricerca

Data science, Computer vision and AI

Biografia

Etibar Vazirov is a PhD candidate in Computer and Control Engineering at Politecnico di Torino, Italy, and a faculty member at ADA University in Azerbaijan. His research lies at the intersection of Graph Neural Networks (GNNs), Explainable Artificial Intelligence (XAI), and data-driven decision support systems.
Traditional data science often treats data as independent rows in a table, ignoring the rich relationships that naturally exist between entities. In contrast, GNNs enable the modelling of these complex interdependencies, allowing learning systems to represent real-world structures more faithfully. Etibar’s research focuses on principled graph construction and explainability — two critical aspects that ensure both accuracy and interpretability in predictive modelling.
He investigates how relational information (such as store proximity, customer region, or behavioural similarity) can be embedded into graph structures to enhance prediction and understanding. In particular, his recent work applies graph-based modelling and explainable AI methods to retail demand forecasting, where nodes represent business entities and edges encode spatial, operational, or similarity-based relations.
Beyond predictive performance, his research aims to develop trustworthy, interpretable systems that assist decision-makers in complex domains. His broader academic interests include trustworthy AI, neural-symbolic reasoning, and the integration of domain knowledge with machine learning for transparent, data-driven insights.

Didattica

Insegnamenti

Corso di laurea magistrale

Ricerca

Gruppi di ricerca

Pubblicazioni

Pubblicazioni più recenti Vedi tutte le pubblicazioni su Porto@Iris