Binary Embedding: Fundamental Limits and Fast Algorithm
Binary embedding is a nonlinear dimension reduction methodology where high dimensional data are embedded into the Hamming cube while preserving the structure of the original space. Specifically, for an arbitrary $N$ distinct points in $\mathbb{S}^{p-1}$, our goal is to encode each point using $m$-dimensional binary strings such that we can reconstruct their geodesic distance up to $\delta$ uniform distortion. Existing binary embedding algorithms either lack theoretical guarantees or suffer from running time $O\big(mp\big)$. We make three contributions: (1) we establish a lower bound that shows any binary embedding oblivious to the set of points requires $m = \Omega(\frac{1}{\delta^2}\log{N})$ bits and a similar lower bound for non-oblivious embeddings into Hamming distance; (2) we propose a novel fast binary embedding algorithm with provably optimal bit complexity $m = O\big(\frac{1}{\delta^2}\log{N}\big)$ and near linear running time $O(p \log p)$ whenever $\log N \ll \delta \sqrt{p}$, with a slightly worse running time for larger $\log N$; (3) we also provide an analytic result about embedding a general set of points $K \subseteq \mathbb{S}^{p-1}$ with even infinite size. Our theoretical findings are supported through experiments on both synthetic and real data sets.
PDF Abstract