An integrated approach for local outlier detection using dynamic minimum spanning tree

Abstract

Detecting outliers in database (as unusual objects) is a big desire. In data mining detection of anomalous pattern in data is more interesting than detecting inliers. Minimum spanning tree based clustering algorithm is capable of detecting clusters with irregular boundaries. In this paper we propose an integrated approach using minimum spanning tree based clustering and Density-based approach for detecting outliers. Outlier detection is an extremely important task in a wide variety of application. The algorithm partition the dataset into optimal number of clusters. Small clusters are then determined and considered as outliers. The rest of the outliers (if any) are then detected in the clusters based on density-based approach. The algorithm uses a new cluster validation criterion based on the geometric property of data partition of the dataset in order to find the proper number of clusters. The algorithm works in two phases. The first phase of the algorithm creates optimal number of clusters, where as the second phase of the algorithm detect outliers. The key feature of our algorithm is it finds noise-free/error-free clusters for a given dataset without using any input parameters.