As a guest user you are not logged in or recognized by your IP address. You have
access to the Front Matter, Abstracts, Author Index, Subject Index and the full
text of Open Access publications.
Clustering analysis is a significant unsupervised machine learning method and plays a more and more important part in multifarious fields. Although many types of clustering methods have been developed for the past few years, most conventional clustering methods can only work with round-shaped clusters and may be affected by the parameter settings and initialization. What's more, the effectiveness of clustering process would be limited by its time and space complexity when the size of the dataset becomes extremely large. In this paper a new and fast grid-based clustering algorithm is proposed which can stably deal with large datasets. In this method, the grid size is automatically determined, and then the local densities at the grid nodes are estimated. Finally, the breadth-first search strategy is used to find the clusters with arbitrary shapes. The method can be successfully applied to various datasets which contain non-spherical clusters with different dimensionalities. Some synthetic datasets and real-world datasets are tested and experiment results show that the proposed method is more effective and more efficient than the conventional methods.
This website uses cookies
We use cookies to provide you with the best possible experience. They also allow us to analyze user behavior in order to constantly improve the website for you. Info about the privacy policy of IOS Press.
This website uses cookies
We use cookies to provide you with the best possible experience. They also allow us to analyze user behavior in order to constantly improve the website for you. Info about the privacy policy of IOS Press.