A Practical Overview of AutoML for Tabular Data @ DevClub.lv #101

Automated Machine Learning (AutoML) became popular in recent years, focusing on cost reduction for building an ML pipeline. The following scientists’ daily tasks can be partially or fully automated: data preparation, feature engineering, model generation, and model estimation. We will briefly overview the existing AutoML solutions and show a practical example based on our winning solution for Sussex-Huawei Locomotion Challenge 2021. In particular, we will discuss how to train a classifier model using Google AutoML Tables for tabular data effectively and with minimal effort.