作者: Dimitrios Gunopulos , Gautam Das , Heikki Mannila , Béla Bollobás
DOI:
关键词:
摘要: Given a pair of nonidentical complex objects, defining (and determining) how similar they are to each other is nontrivial problem. In data mining applications, one frequently needs determine the similarity between two time series. We analyze model time-series that allows outliers, different scaling functions, and variable sampling rates. present several deterministic randomized algorithms for computing this notion similarity. The based on tools methods from computational geometry. particular, we use properties families well-separated geometric sets. algorithm has provably good performance also works extremely efficiently in practice.