Foundations of causal inference and its impacts on machine learning webinar

Опубликовано 29 марта 2021, 18:19

Many key data science tasks are about decision-making. They require understanding the causes of an event and how to take action to improve future outcomes. Machine learning (ML) models rely on correlational patterns to predict the answer to a question but often fail at these decision-making tasks, as the very decisions and actions they drive change the patterns they rely on. Causal inference methods, in contrast, are designed to rely on patterns generated by stable and robust causal mechanisms, even as decisions and actions change. With insights gained from causal methods, the new, growing field of causal machine learning promises to address fundamental ML challenges in generalizability, interpretability, bias, and privacy.

In this webinar, join Microsoft researchers Amit Sharma and Emre Kıcıman to learn about the fundamentals of causal inference. You will learn how a target question of cause and effect can be captured in a formal graphical model and answered systematically using available data. The researchers will introduce a four-step causal modeling framework for analyzing decision-making tasks and walk-through code examples using the DoWhy Python library that implements the framework. You will also discover how causal methods can be useful to improve ML models in terms of their generalizability, explainability, fairness, and robustness.

Together, you’ll explore:

■ Why causal reasoning is necessary for decision-making
■ The difference between a prediction and a decision-making task
■ How the DoWhy library can help you conduct a robust causal inference analysis by translating domain knowledge to a causal graph and validating the graph using available data
■ The connections between causal inference and the challenges of modern ML models

Amit Sharma is a Senior Researcher at Microsoft Research India. His work bridges causal inference techniques with data mining and machine learning, with the goal of making machine learning models generalize better, be explainable and avoid hidden biases. To this end, Amit has co-led the development of the open-source Microsoft DoWhy library for causal inference and DiCE library for counterfactual explanations.

Emre Kiciman is a Senior Principal Researcher at Microsoft Research at Redmond. His research is motivated by decision-making tasks in a variety of societally critical domains and includes research on causal machine learning and the security of AI systems. Emre is a co-author of the DoWhy library for causal inference.

𝗥𝗲𝘀𝗼𝘂𝗿𝗰𝗲 𝗹𝗶𝘀𝘁:

■ Foundations of causal inference and its impacts on machine learning (presentation slides)
■ DoWhy: Causal Reasoning for Designing and Evaluating Interventions (project page) microsoft.com/en-us/research/p...
■ Causal Reasoning: Fundamentals and Machine Learning Applications (book excerpt): causalinference.gitlab.io
■ Causality and Machine Learning at Microsoft (publications): microsoft.com/en-us/research/g...
■ DoWhy: A library for causal inference (GitHub) github.com/Microsoft/dowhy
■ DoWhy: An end-to-end library for causal inference (paper): arxiv.org/abs/2011.04216
■ DoWhy – A library for causal inference (blog) - microsoft.com/en-us/research/b...

*This on-demand webinar features a previously recorded Q&A session and open captioning.

This webinar originally aired on December 3, 2020

Explore more Microsoft Research webinars: aka.ms/msrwebinars