This vacation project will look at robust variance adjustments to the random effects Conway-Maxwell-Poisson model, so that resulting parameter inferences can be adaptive to possible misspecifications of the variance-covariance structure.
The University of Queensland
Wilson is in his fourth year completing a dual degree of Bachelor of Mathematics (Statistics) / Commerce (Finance) at The University of Queensland (UQ). The classical statistics using probabilistic data models have been his interests to develop more useful findings of statistical significance variables through observational and experimental studies, especially in the field of psychology, health, and financial-economic. Thus, a lot of established models would be assessed against test statistics to give insightful results. Also, his interests would encompass modern statistics using predictive performance as a measure of “success” because of its utilisation of many “black-box” algorithms that could give a low-error rate in predicting new data. This was particularly useful in some daily areas, for example predicting spending behavior for certain demographics, or smartphone face recognition for password utilisation. The AMSI Vacation Research Scholarship will give him a stepping stone for developing a further academic career, especially into an honours and/or PhD program. Besides, it will help him applying concepts in practice and communicating ideas to the public in an efficient and effective way.