Data fusion approach for enhanced anomaly detection

Abstract

Anomaly detection is very sensitive for the data because, the feature vector selection is a very influential aspect in the anomaly detection rate and performance of the system. In this paper, we are trying to revise the dataset based on the rough genetic approach. This method improves the quality of the dataset based on the selection of valid input records to enhance the anomaly detection rate. We used rough sets for pre-processing the data and dimensionality reductions. Genetic algorithm is used to select proper feature vectors based on the fitness. The fusion of the soft computing techniques improves the data quality and reduces dimensionality. Empirical results prove that it improves detection rate as well as detection speed.