Inference mobility preferences and Optimization via Graph signal processing
Tesi esterna in azienda
Parole chiave COLLABORATIVE FILTERING, GRAPH SIGNAL PROCESSING,
Riferimenti FABRIZIO DABBENE
Descrizione The problem of inferring user scores/preferences of different items from partial observations arises in many
problems data are related by dependency or similarity relations and this information can be exploited for a better prediction.
For example, there exist relationships between users (such as their age, gender, hobbies, education, etc). This information can be taken advantage of, since people sharing the same tastes for a class of items are likely to rate them similarly. Graph theory provides an attractive and innovative method for the mathematical modeling of this correlation.
We are interested in (a) applying graph-based techniques to model and possibly predict the users’ inclinations, opinions and beliefs on all aspects of a vehicle-sharing system; (b) proposing quantitative dynamical models to describe the formation of users’ opinion and beliefs and identify the minimum set of “influential individuals” who can guide the users’ choices and to boost acceptance of sharing solutions.
Conoscenze richieste Basic mathematical modeling, Basic signal processing, Convex optimization, Basic coding skills (MATLAB is
Note The thesis will be carried out under the supervision of the staff of the System and Modeling Control Group at the Institute of Electronics, Computer and Telecommunication Engineering of National Research Council of Italy (IEIIT-CNR), Prof. Mara Tanelli of Politecnico di Milano (DEIB) and other members of the MoVE group of the DEIB (www.move.deib.polimi.it).
Scadenza validita proposta 20/11/2019 PROPONI LA TUA CANDIDATURA