Talk title: An overview of b-bit minwise hashing, and two proofs for the main theorem
In 2010, Ping Li and Arnd Christian Konig established the framework for b-bit minwise hashing, which is an efficient method of estimating set similarities R. They came up with an approximate formula to estimate the main probability result, and gave an elegant proof of it. In today's talk, I will give a brief overview of b-bit minwise hashing, Ping's and Konig's proof, and yet another proof of the main result, which may lead to a more accurate formula (for small D). Joint work with Ping Li.