Pre- Conference Workshop
A Short course in Automatic Model Selection | 12 April
This one day, live and in-person course will introduce attendees to the theory and practice of automatic model selection for econometric modelling of economic and climate variables in a non-stationary world. It covers the modelling methodology, implementation, practice, and evaluation of econometric models. The course will focus on practical, hands-on learning, as well as explanatory sessions.
This hands - on course will run live and In-Person at Etc Venues, 50 - 52 Chancery Lane, London, UK.
This course is aimed at professionals and advanced students who wish to learn more about automatic modelling selection.
The content provides background information that is relevant for the Dynamic Econometrics conference.
By the end of this course participants should:
understand the challenges involved when working with time-series data.
Be confident handling evolving time series exhibiting trends, outliers, and sudden shifts.
Develop skills in evaluating and selecting empirical econometric models.
Upon Completion of this course, all participants will receive a certificate of completion, free entry for one to the 2023 Dynamic Econometrics Conference and a months trial use of OxMetrics 9.
Econometric Introduction | 11 - 12 AM
Economic time series exhibit evolving stochastic trends and sudden distributional shifts. We use historical UK data from over a century to illustrate trends and breaks, and how they may be handled. The econometric tools are introduced, as well as OxMetrics, the software used for estimation and Monte Carlo experiments.
Introduction to Model Selection | 12 - 1 PM
We discuss the Theory of Reduction, revealing the steps that lead us to our empirical model, as well as approaches to automatic model selection, and their properties. We discus Autometrics, the machine learning tool that we use for model discovery
Automatic Model Selection and Autometrics | 2 - 3 PM
We consider several approaches to automatic model selection, and their properties. We introduce Autometrics, the machine learning tool that we use for model discovery.
Outliers and Breaks - Saturation Estimators | 3 - 4 PM
We describe a computational approach that can be used in a setting that requires the implementation of indicator and step saturation estimators (e.g practical problems such as pandemics, energy crises and financial crises). Monte Carlo and practical examples are used to illustrate its performance.