Skip to content
nonconvexcard

The random-hyperplane collision probability

ProofIntermediate⏱ 30 minRAGlshsimhashanglerotational invariance

Problem

Prove P(sign⁡⟨w,u⟩≠sign⁡⟨w,v⟩)=θ(u,v)/π\P(\sign\langle w,u\rangle\ne\sign\langle w,v\rangle) = \theta(u,v)/\pi for w∼N(0,Id)w\sim\Ncal(0,I_d), and deduce the variance of a bb-bit SimHash estimate of the angle.

Write a complete argument. Every inequality you use should be one you could state precisely.

Solution

Opening this before you have a real attempt on paper is the fastest way to feel like you understand something you cannot reproduce.

Related problems

All problems
SolvedProblemFieldTypeDifficultyTime
Product quantisation budget arithmeticRAGNumericalBeginner30 min
Recall composes multiplicativelyRAGNumericalBeginner15 min
Implement SimHash LSH and measure recall runRAGCodingAdvanced3 hours
Why space-partitioning trees fail in high dimensionRAGProofAdvanced30 min
The LSH amplification exponent abyssRAGProofGraduate1 hour