Machine learning techniques for collective behavior identification
keywords BIG DATA ANALYSIS, COMMUNICATIONS NETWORKS, DARKNET, DATA SCIENCE, GRAPHS, MARKOV CHAINS, MODELLING, SOCIAL NETWORKS, STATISTICS
Reference persons MARCO MELLIA, LUCA VASSIO
External reference persons Idilio Drago (UNITO)
Research Groups SmartData@PoliTO, Telecommunication Networks Group
Thesis type DATA ANALYSIS, MODELLING, NUMERICAL APPROXIMATION
Description In complex systems, there are interactions between entities and we want to characterize and extract the salient types of these interactions. The main applications we are interested to are online social networks like Facebook and Instagram or communication network, like darknets listening to incoming Internet traffic. The student will:
• Develop machine learning methodologies to identify coordinated events .
• Model the events as a graph and extract emerging patterns.
• Make the methodological approach scalable with smart approximations.
• Apply these methods to real networks (darknets, social networks), for which we already collected a lot of data
Required skills Modelling, numerical methods, data analysis, python, possibly Big Data (pyspark)
Notes Contact firstname.lastname@example.org
Deadline 16/03/2022 PROPONI LA TUA CANDIDATURA