Machine Learning in Asset Management using Density Networks

Event Date: 

Wednesday, April 3, 2024 - 3:30pm to 4:45pm

Event Location: 

  • HSSB 1174

Event Price: 


Event Contact: 

Short Bio: 

Jennifer Chan obtained her PhD from the University of New South Wales in 1997. She joined The University of Hong Kong in 1996 and University of Sydney in 2006. She was promoted to Associate Professor in 2017. Her research interest includes Generalised Linear Models, Bayesian Robustness, volatility and covariance modelling, loss reserving and machine learning using neural networks.

  • Department Seminar

This research aims to transform investment strategies through the application of advanced machine learning via neural networks and statistical modeling, with a focus on creating interpretable asset price prediction models and optimizing portfolio strategies. This project advances asset price forecasting through mixture models, enhancing interpretability and providing deeper insights into asset return dynamics. It bridges the gap by intertwining the asset price modeling with corresponding investment strategies, thereby enhancing the risk-adjusted returns of asset investments. This project is extendable to the domain of mean-variance portfolio optimization by employing multivariate mixture models for portfolio returns. The multivariate models facilitate precise estimation of the portfolio's covariance matrices and expected mean returns, effectively addressing the prevalent gaps in portfolio return estimation. We endeavor to provide actionable tools for financial analysis and decision-making, poised to make a significant impact on the investment industry by fostering more informed investment decisions and enhancing portfolio management.