Time Series Analysis and Machine Learning - Q4

Participation Prerequisites

Familiarity with the topics covered in Statistics I and II as well as with the material covered in Causal Inference & Reasoning is assumed. A good summary is provided in Stock & Watson (2020, Chs. 2–7).

Course Content

This course is centered around the problem of predicting time series. Applications are taken from finance, macroeconomics and the earth sciences. Classical methods for time series modelling such as autoregressions are introduced and discussed in detail. This is followed by a treatment of advanced techniques such as unit root processes, structural breaks and GARCH models. Finally, machine learning is covered to cater for present-day massive datasets. The empirical analyses are implemented in RStudio, the most popular data science software environment, and in RMarkdown, the prime language for producing replicable research.

Intended Learning Outcomes and Competencies

By the end of this course, students will be familiar with modern econometric and machine learning techniques and will be able to apply them to real world data sets using state-of-the-art software. Students will have acquired a sound theoretical mindset for time series prediction. They will have developed programming skills for conducting replicable empirical work. This proficiency will prove indispensable for their Bachelor’s thesis, for a Master’s degree or for data science projects in industry.

Instruction Type

In-class.

Form of Examination

Literature

Stock, J. H. & M. W. Watson (2020). Introduction to Econometrics. 4th ed. Boston: Pearson. ISBN: 9781292264455. URL: https://media.pearsoncmg.com/intl/ge/2019/cws/ge_stock_econometrics_4/.