Abstract: Based on key point technology, a new method for time series cluster was proposed. hhe key points for each time series were first found, and then the complex network was constructed by calculating the similarity between key point series after they were ectuidimensional. At last, the clustering time series were implemented by partitioning the complex network into communities. hhe experimental results show that the dimensions of time series and the consumplion of computing time can be effectively reduced by the proposal. Furthermore, the desired cluster result is obtained when applying this method to cluster some practical data.

Key words: Time series, Reduction dimension, Key point, Complex network, Cluster