State space models in credit and operational risk management

This project targets development of applied methods and practical solutions to risk management problems where only partial observation of a system is possible. Such settings are commonplace in financial and other context but can be challenging to address due to a limited number of production-grade ready-to-use solutions. The scientific component of the project employs approaches from a quickly developing and active area in machine learning. More extensive use of these approaches by Canadian banking institutions will lead to a more robust financial system and better service. The focal application areas in this project target the improvement of the value proposition to bank's customers, accurate prediction of credit losses, and better sales outcomes in multi-product salesforces of financial services.

Amir Emami Gohari;Alireza Farhang Doost;Linh Dieu Thi Dang;Vikram Dhingra
Faculty Supervisor: 
Yuri Levin;Mikhail Nediak;Jue Wang
Partner University: