Machine Learning Forecasting Techniques For Company Financial Fundamentals In Long-Term Value Investing

This project is addressing a deficit in the long-term valuation space for robust mathematics. Current techniques rely on very basic statistics and manually constructed spreadsheet models. This represents significant operational risk to investment firms. This project will seek to, using machine learning, develop accurate forecasting models for the financial fundamentals of companies in order to help address this risk. These models, mainly using financial time series forecasting techniques, will extend existing forecasting algorithms for companies’ financial fundamentals to improve upon the accuracy of pre-existing techniques. Complementary to this will be further development on research into making these machine learning algorithms interpretable by non-technical users.

Faculty Supervisor:

Roger Grosse

Student:

Partner:

Valsys

Discipline:

Computer science

Sector:

Finance and Insurance

University:

University of Toronto

Program: