Μεθοδολογίες μηχανικής μάθησης για την πρόγνωση πιστωτικού κινδύνου
Machine learning methodologies for credit risk prediction
Based on the evolution of technology in recent decades in terms of machine learning, financial institutions are given the opportunity to use it to their advantage. A modern and widespread problem in this type of business is credit risk. Computer science is therefore called upon to provide a solution to a problem that seemed impossible to solve. With the extraction of data whose volume seems prohibitive for human management, but also with special models of machine learning, it was possible to increase the probability of avoiding credit risk. In this paper, therefore, all these concepts are analyzed and a special problem is solved for a company that grants loans and tries to achieve as easy and immediate return of funds as possible.