Home

Industry has adopted controlled experiments for software and product development purposes. Microsoft’s Bing development team alone, runs over 10,000 experiments a year. Startups that endorse controlled experiments tend to outperform those that do not.

However, establishing a successful experimentation practice in a company requires implementing a computationally powerful platform that combines the best of science and engineering. Scientific methods in particular, amid “big-data” have to overcome the data-rich and information poor regime, as well as ensure inference validity regardless of the business and engineering constraints.

In this tutorial, the participants will familiarize with the inference techniques that have been developed to tackle these causal inference challenges. Participants will get hands-on experience, applying techniques at the intersection of causal inference and machine learning. This tutorial will provide ECML PKDD participants with thorough, in-depth, and hands-on experience with causal inference and machine learning techniques used and developed in industry and academia - stimulating further scientific advancements and collaboration in the community.