MSc Thesis Proposal Announcement of Ahmed Shafeek:"Tree-Based Approaches to Predicting Borrowers’ Financial Performance "

Date:                  Thursday February 3, 2022 

Time:                  9:00am – 10:00am 

Meeting URL:     https://us06web.zoom.us/j/87540713915?from=addon 

Passcode:          If interested in attending this event, contact the Graduate Secretary at csgradinfo@uwindsor.ca with sufficient notice before the event to obtain                                  the  passcode.

 

The literature review we conducted pointed to that decision trees are the most successful approach in modeling financial performance and credit scoring. Comparative studies showed tree-based methods’ superiority over deep learning and other neural networks for tabular and categorical datasets. We will investigate applying Tree-Based Methods to find the feature importance and accurately predict financial performance. We chose CatBoost platform due to its efficiency, simplicity, and accuracy. We were given access to a proprietary US-based dataset, to mine and analyze, on the condition of namelessness. Profitability, risk, and bottom lines are all non-factors for this specific group of people. Predictions will not be used to refuse loans higher risk population, but rather to provide them with services and training that can better prepare them for success; a humanitarian approach to a financial problem. 

 

Keywords: Financial Performance Prediction, Credit Scoring, Delinquency Rate, Decision Trees, Tree-based Methods, CatBoost 

 

Internal Reader: Dr. Ziad Kobti    

External Reader: Dr. Hoda ElMaraghy      

Advisor: Dr. Alioune Ngom