
… why we see a huge difference in model performance between offline evaluation and final production. It is high time we stop treating ML as a “black box” and give importance to model interpretation while improving model accuracy. This will also help us in identifying data biases. In this part, we will see how catboost helps us …
…d machine learning, that also models objects and processes around us in a totally different manner. While mathematicians create models based on their expertise and understanding of the world, machine learning algorithms describe the world in some hidden way, not fully understandable, but in most of the cases even more accurate than mathematical models developed by human experts. However, in a lot of applications (as healthcare, finance, military), we need clean and interpretable decisions, that machine learning algorithms, in particular, deep learning models are not designed to provide.
…e have a new caste system in the United States. Our society is now composed of two types of people. The first type are early adopters of technology who were privileged enough to have access to a technology education at a young age and who have accelerated their own personal and professional development. The second type of person is the vast majority of the citizens of the United States of America.