Fast nearest-neighbor search in dissimilarity spaces
András Faragó, Tamás Linder, Gábor Lugosi · IEEE Transactions on Pattern Analysis and Machine Intelligence · 1993
A fast nearest-neighbor algorithm is presented. It works in general spaces in which the known cell techniques cannot be implemented for various reasons, such as the absence of coordinate structure or high dimensionality. The central idea has already appeared several times in the literature with extensive computer simulation results. An exact probabilistic analysis of this family of algorithms that proves its O(1) asymptotic average complexity measured in the number of dissimilarity calculations is presented.>