Detecting Suspicious Companies
By Alejandro Beltran Project description Complicated company ownership structures are often set up to obfuscate the identification of beneficial owners. These strategies are considered to be on the fringe of legitimacy and legality, so detecting them is important to regulators and society in general. We combine economic theory, network analysis, and machine learning to unearth the patterns that such behaviours leave behind in large-scale administrative datasets. This allows us to identify clusters of companies that specialise in such strategies, and to formulate models about their potential adaptation to future legal reforms....