IBORA:an Improved Efficient Detection of Boundary Points

Proposed a new boundary detecting algorithm called IBORA(an Improved BOundaRy points detection based on Angle)based on the problem of the current low efficient boundary detection and the scope of the parameter pruning is not easy determined in noisy datasets.The algorithm detects boundary points based on the two features:First,A bigger angle area of being constituted by two vectors no longer contains any point,and the two vectors have been made up of by the boundary point with the other points in Eps neighborhood.Second,Outliers distribute sparsely in the space.Experimental results indicate that IBORA can detect boundary points in noisy dataset containing different shapes and sizes clusters effectively,and has higher efficiency.