Outlier detection algorithm based on SOM neural network for spatial series dataset
Yaxian Liu, Hui Lu · 2018
Outlier detection is an important branch of data mining which has been applied in different fields. Facing the multidimensional spatial series dataset containing both isolated and assembled outliers, many existing methods become unsatisfactory or even inapplicable. In this paper, we propose an outlier detection algorithm based on SOM (Self-Organizing Maps) neural network for the spatial series dataset. Firstly, we introduce the principle of the clustering algorithm based on SOM neural network. Secondly, the outlier detection strategy is designed according to the topological distribution of neurons. Finally, in order to verify the effectiveness and reliability of the proposed algorithm, several experiments are performed in this paper. The simulation illustrates that the proposed algorithm based on SOM neural network is very effective and reliable for spatial series dataset.