In recent years machine learning and causal inference have both seen important advances, especially through a dramatic expansion of their theoretical and practical domains. Machine learning has focused on ultra high-dimensional models and scalable stochastic algorithms, whereas causal inference has been guiding policy in complex domains involving economics, social and health sciences, and business. Through such advances a powerful cross-pollination has emerged as a new set of methodologies promising to deliver robust data analysis than each field could individually -- some examples include concepts such as doubly-robust methods, targeted learning, double machine learning, causal trees, all of which have recently been introduced.

This workshop is aimed at facilitating more interactions between researchers in machine learning and causal inference. In particular, it is an opportunity to bring together highly technical individuals who are strongly motivated by the practical importance and real-world impact of their work. Cultivating such interactions will lead to the development of theory, methodology, and - most importantly - practical tools, that better target causal questions across different domains.

In particular, we will highlight theory, algorithms and applications on automatic decision making systems, such as recommendation engines, medical decision systems and self-driving cars, as both producers and users of data. The challenge here is the feedback between learning from data and then taking actions that may affect what data will be made available for future learning. Learning algorithms have to reason about how changes to the system will affect future data, giving rise to challenging counterfactual and causal reasoning issues that the learning algorithm has to account for. Modern and scalable policy learning algorithms also require operating with non-experimental data, such as logged user interaction data where users click ads suggested by recommender systems trained on historical user clicks.

To further bring the community together around the use of such interaction data, this workshop will host a Kaggle challenge problem based on the first real-world dataset of logged contextual bandit feedback with non-uniform action-selection propensities. The dataset consists of several gigabytes of data from an ad placement system, which we have processed into multiple well-defined learning problems of increasing complexity, feedback signal, and context. Participants in the challenge problem will be able to discuss their results at the workshop.