A new algorithm for similarity search in distributions achieves sublinear sample complexity & query time. This opens up opportunities for efficient data analysis and improved ML applications: http://
bit.ly/3PuiQ2R

→ View original post on X — @mit_csail