A new algorithm for high-dimensional outlier detection based on constrained particle swarm intelligence

In this paper we present an algorithm for outlier detection in high-dimensional spaces based on constrained particle swarm optimization techniques. The concept of outliers is defined as sparsely populated patterns in lower dimensional subspaces. The search for best abnormally sparse subspaces is done by an innovative use of particle swarm optimization methods with a specifically designed particle coding and conversion strategy as well as some dimensionality-preserving updating techniques. Experimental results show that the proposed algorithm is feasible and effective for high-dimensional outlier detection problems.