preface

The first time causal inference caught my attention was when I read a paper in 2016. It talked about causality, and I understood absolutely none of it. Curious about this exotic topic, I read Judea Pearl’s works. Initially, I didn’t expect much could be done with mathematical and statistical tools to model causality. But as I kept reading, I realized a lot can be done with causality from an applied perspective.

Back in 2016, I had been working as a data scientist for a few years. I enjoyed learning machine learning techniques; it was a new world, and my technical background was an advantage. Machine learning opened the doors to a variety of industries and companies, allowing me to enjoy my work and be productive.

Between 2016 and 2018, I began to sense that much of machine learning’s progress was achieved through trial and error, without a deep understanding of its inner workings, and deep learning played a major role in this approach. The focus was on computing power and programming rather than on modeling the world. I was not against this approach, but it didn’t fully satisfy me. At the same time, I was delving into causal inference. Each day, I found myself more aligned with its goals—uncovering the “why"—and its methods, which place a greater emphasis on statistics and mathematics while still incorporating programming.