Machine learning for economics

This course focuses on supervised and unsupervised machine learning methods, that are very useful when studying economic phenomena. In economics, forecasting is frequently a main goal and thus, supervised methods are developed because they help in facing a prediction task (regression techniques for continuous target variables; classification tools for discrete target variables). On the other hand, nowadays large datasets are commonly available. For this reason, in this course the unsupervised methods are presented, as they are helpful for dimension data-reduction and for discovering underlying data-structures.
This course enables students to learn which specific statistical tool should be applied for a particular goal. Moreover, the course aims at teaching a statistical software to analyse real datasets. Data applications and analyses allow students to develop problem solving and programming skills.

At the end of the course students will be able to apply machine learning techniques and algorithms in economic settings. Specifically, they will become familiar with both unsupervised and supervised methods. In the unsupervised framework students will be able to perform principal component analysis (useful for dimensionality reduction) and cluster analysis (useful to discover underlying structures in the data). In the supervised framework students will be able to apply advanced regression and classification techniques to face any prediction task. Students will acquire independence in studying statistical and machine learning tools and will be able to solve practical problems in economic data analysis.