Member-only story
Will AutoML Be the End of Data Scientists?
AutoML is exploding in popularity. Here’s how that changes things.
Background
In 2012, an arXiv report on Auto-WEKA was released, describing an automated approach to selecting a machine learning algorithm, features, and hyperparameters, in the hopes that it would “help non-expert users” in the field.
More recently, AutoML has exploded in popularity, with all the big tech players entering the space.
In April 2016, Facebook declared that AutoML was a “backbone” of its AI. In January 2018, Google announced Cloud AutoML. In August 2018, Salesforce open-sourced its Einstein AutoML library. A month later, Microsoft introduced AutoML to its Azure AI platform. Earlier this year, Amazon introduced AutoGluon, another open-source AutoML toolkit.