Improving Likelihood Estimators: Theory and Applications to Analyzing Productivity and Efficiency and Forecasting of Probability of Economic Recession (2013–2018)

Parametric maximum likelihood method is one of the most popular methods of statistical analysis. Its main weakness is requirement of parametric assumptions. Non-parametric or local maximum likelihood method has been developed but only for the case of continuous explanatory variables and so it has rarely been used because data sets often contain discrete variables. In this project we aim to expand the theory of this method to allow for discrete regressors and for time series contexts and will use it to unveil patterns of economic growth, productivity & efficiency of countries, and to forecast probability of an advent of economic recession. Our work will generalize existing methods and empower applied researchers with more reliable tools.
Grant type:
ARC Discovery Projects
Funded by:
Australian Research Council