| KEYWORD |
Exit prediction for Privately Held Companies
Parole chiave CLASSIFIERS, FINANCE, MACHINE LEARNING, PREDICTION METHODS
Riferimenti GIUSEPPE CARLO CALAFIORE
Gruppi di ricerca SYSTEMS AND DATA SCIENCE - SDS
Tipo tesi RICERCA APPLICATA
Descrizione Predicting the exit (e.g. IPO, bankrupt, acquisition, etc.) of privately held companies is a current and relevant problem for investment firms, whose difficulty stems from the lack of reliable and publicly available data.
In this thesis we shall study exit predictor models based on qualitative data. Methodologies will include Logistic Regression models, Random Forest models, Support Vector Machine models, as well as Survival models. Experimental tests will be conducted on available data from a Thomson Eikon repository.
Conoscenze richieste Machine learning, optimization, basics of corporate finance.
Scadenza validita proposta 10/12/2019
PROPONI LA TUA CANDIDATURA