Kaggle, a website that hosts public competitions on machine learning tasks, announced Tuesday that it now has over 1 million users a little more than seven years after it launched.
It’s a sign of accelerating interest in the artificial intelligence and machine learning field, since Kaggle competitions give data scientists a way to test techniques on real-world tasks with the chance to compete for prize money. Competitions work by providing teams with a data set and then evaluating how well the models they create succeed at the task laid out.
The company’s first competition began April 7, 2010, with 22 teams attempting to predict votes on the Eurovision music competition. Since then, it has become a key part of the data science ecosystem. Tasks currently available include a competition sponsored by real estate site Zillow, which is offering a prize pool of $1.2 million for teams that can enhance its home price estimation algorithm.
Kaggle has seen the rise of several cutting edge machine learning techniques, including decision trees and deep neural networks. One of the site’s key benefits to the data science community is that it facilitates sharing of information about how to solve problems, which can help drive the field forward.
The news comes a few months after Google announced its acquisition of Kaggle as part of the company’s focus on artificial intelligence.