Application of Neural Network and Cluster Analysis in Anomaly Detection

This paper, taking the applying data of financial subsidy for agriculture development as an example, makes use of cluster analysis to set anomalous points detection data stream on the basis of data source and finds the most possible fields which may cause abnormal records; then uses neural network and cluster analysis method respectively to analyze and compare the abnormal records for finding out the data likely carrying the humbug, which could be consulted by management department when making decisions.