Melvern Leung is a part of the Risk Model Validation team in Bendigo and Adelaide Bank.
PhD in Econometrics and Business Statistics, 2020
Monash University
Honours in Mathematics (Statistics), 2015
Monash University
BSc in Pharmacology and Statistics, 2014
Monash University
Helped in the development of the Risk Model validation standards. These include definitions of what models need to be validated, how validation will take place and what will be the difference between periodic and development validations.
Periodic and development validations of Credit Risk Models such as PD, Application Scorecard, LGD and EAD models.
Validation of market risk models, of which included was the Value-At-Risk and backtesting models.
Lead a group of consultants and senior consultants in coding an automated timeseries forecasting package and utilizing Recursive Feature Elimination.
Helped an insurance company develop a macro-economic forecasting model to analyse the effects of climate change on the probability of default and actuarial credit spreads across various industries.
Developed multiple dynamic microsimulation model for various policy and scenario analysis.
Lead the development of an R shiny program to undergo geospatial data analysis with a focus on data visualization.
Utilized a variety of programming languages to help clients solve their data analytics problem within their cloud environment (Python, R, TeradataSQL, and PowerBI).
Trained graduates and consultants the coding skillsets needed to meet client deliverables
Managed and presented proposals to clients and helped ensure the engagement milestones were on track
Responsibilities include:
Seminar speech at The Longevity 12 Conference in Chicago, US, 2016
Seminar speech at The Quantitative Finance and Risk Analysis Conference in Kos Island, Greece, 2019
This project aims to apply and adapt the latest machine learning techniques to enable companies to utilise their existing customer data to reveal purchase motivations, product preferences, and responsiveness to marketing communications for each single customer.
Bayesian Neural Classifiers in FOREX trading
My objective is to apply new state-of-the-art Bayesian econometric tools to the area of longevity risk management.
In this paper we provide an insight to the study of several pricing approaches for longevity instruments that have been proposed in the literature. To account for parameter uncertainty in mortality forecasts and longevity instruments pricing, our analysis hinges on a Bayesian state-space mortality model.
