Causal models have many advantages, including the ability to reason about the effects of interventions, as well as the results of different scenarios or counterfactuals, facilitating informed decision-making and problem-solving. The traditional approach for building causal models is by conducting experiments, however these are often infeasible, unethical or too expensive.

Recently there has been a lot of interest in the scientific community to learn causal models from observational data across various sectors, including healthcare, economics, social sciences, and artificial intelligence. Causal discovery enhances predictive modeling by incorporating causal relationships, reducing reliance on mere correlations, but this is a great challenge, as just from observations is not possible, in general, to define a unique causal model. One of the areas that has recently benefited from causal discovery is the field of time series. The causal analysis enables the identification of interventions that lead to desired outcomes. In fields like healthcare, it informs medical decisions and improves patient care by uncovering the causal links between treatments and outcomes.

The main objective of this tutorial is to introduce to the audience to the area of causal discovery, including the fundamentals and software tools. Initially, we will introduce graphical causal models, including the knowledge representation and inference techniques. Will present a comprehensive introduction to causal discovery methods equipping participants with the knowledge and skills to understand the fundamental concepts of causality and causal inference. Finally, we will have a practical session with some software tools for causal discovery in time series.

Target Audience

The tutorial is designed for computer scientist and engineers interested in machine learning and causal modelling. As the tutorial starts from the basics, it can accommodate participants with varying levels of prior knowledge, from beginners seeking an introduction to causal discovery to practitioners looking to deepen their expertise.

Pre-requisites

Basic knowledge of computer science is assumed, as well as programming experience (Python). Some prior knowledge on probability, machine learning and Bayesian networks will be useful, but not required.

Technical requirements

Attendees should bring their own laptop and internet connection for the practical part of the tutorial. It is recommended to use Google Collab to follow the practical steps in time series analysis; the material is available at https://colab.research.google.com/drive/19notnYilbswEfWwjBJOzCruaduH5b3qq?usp=sharing. Alternatively, you can install a virtual environment to work locally. The necessary requirements are available here: https://github.com/Cadismx/CaDis2023/tree/5989cd668a2fec9040f926a4a66711cff9ed63c8/Tools%20for%20Causal%20Discovery%20in%20Time%20Series