Nearest neighbor search algorithm

Abstract EN

A fundamental activity common to image processing, pattern recognition, and clustering algorithm involves searching set of n , k-dimensional data for one which is nearest to a given target data with respect to distance function.

Our goal is to find search algorithms with are full-search equivalent -which is resulting match as a good as we could obtain if we were to search the set exhausting.

1- Aim of the work .

We propose a framework made up of three components, namely l.

A technique for obtaining a good initial match.

2.

An inexpensive method for determining whether the current match is a full- search equivalent match.

3.

An effective technique for improving the current match.

Our approach is to consider a good solution for component in order to find an algorithm, which balances the overall complexity of the search.