Introductory description

Analyses in all fields of Economics nowadays make frequent use of large and detailed datasets ("big data"). Their increased availability opens up many opportunities for applied research, as well as new challenges on how to handle, process, and extract meaningful conclusions from the data. This module provides an overview of recent developments in econometric methods tailored to handle such large datasets, both in supervised and unsupervised methods, including machine-learning techniques Further discussion is provided on the applicability of those methods relative to other Econometric methods.

Module aims

Outline syllabus

This is an indicative module outline only to give an indication of the sort of topics that may be covered. Actual sessions held may differ.

• Principal Components
• Lasso, Adaptive Lasso, Elastic Net, Penalized Logistic Regression
• Random Forest, Regression trees
• Neural Networks
• Topic modelling, text analysis

Learning outcomes

By the end of the module, students should be able to:

• Subject Knowledge and Understanding: Be able to use a variety of modern data-science methods to solve economic questions. The teaching and learning methods that enable students to achieve this learning outcome are: Lectures, seminars, independent study. The summative assessment methods that measure the achievement of this learning outcome are: Exam
• Subject Knowledge and Understanding: Be able to use R to process data and apply data-science methods. The teaching and learning methods that enable students to achieve this learning outcome are: Lectures, seminars, independent study. The summative assessment methods that measure the achievement of this learning outcome are: Exam
• Subject Knowledge and Understanding: Understand under which conditions each method applies and be able to adapt their strategy to the problem studied. The teaching and learning methods that enable students to achieve this learning outcome are: Lectures, seminars, independent study. The summative assessment methods that measure the achievement of this learning outcome are: Exam
• Subject Knowledge and Understanding: Be able to use methods for both predictive and causal purposes. The teaching and learning methods that enable students to achieve this learning outcome are: Lectures, seminars, independent study. The summative assessment methods that measure the achievement of this learning outcome are: Exam
• Subject Knowledge and Understanding: Be able to understand, distinguish, and communicate the differences between correlational and causal analysis in the context of big data and machine learning methods. The teaching and learning methods that enable students to achieve this learning outcome are: Lectures, seminars, independent study. The summative assessment methods that measure the achievement of this learning outcome are: Exam

Indicative reading list

Reading lists can be found in Talis

Interdisciplinary

Module covers contents that are relevant to also disciplines including Computer Sciences and Statistics. Applications of the method might involve other disciplines.

International

The methods presented in the module may be applied to various contexts; in fact, students are encouraged to bring applications and research questions that are pertinent to their international backgrounds.

Subject specific skills

• Develop a critical understanding of statistical methods for large datasets, including both supervised and unsupervised methods.
• Be able to apply such methods in the open-source language R
• Enhance the capacity to conduct economic analyses autonomously
• Be able to understand, distinguish, and communicate the differences between correlational and causal analysis
  in the context of big data and machine learning methods

Transferable skills

• Develop a critical understanding of statistical methods for large datasets, including both supervised and unsupervised methods.
• Be able to apply such methods in the open-source language R
• Enhance the capacity to conduct economic analyses autonomously
• Be able to understand, distinguish, and communicate the differences between correlational and causal analysis
  in the context of big data and machine learning methods