Locality-sensitive hashing (LSH) allows for fast retrieval of similar objects from an index - orders of magnitude faster than simple search at the cost of some additional computation and some false positives/negatives. In the last post I introduced LSH for angular distance. In this one I will tell you how you can fine-tune it to get the expected results.

Locality-sensitive hashing (LSH) is an important group of techniques which can be used to speed up vastly the task of finding similar sets or vectors.