Deep Learning Based Cost Constraint Algorithm for Intrusion Detection Feature Extraction

The defense performance of intrusion detection system is often affected by class unbalance data. In order to automatically extract data features of scarce classes to improve the accuracy of intrusion detection systems in identifying unknown network attacks, a cost constraint algorithm is proposed. Firstly, build a deep neural network based on a stacked autoencoder, and add sparse constraints on the neurons in the hidden layer. Secondly, optimize the cost objective function by generating a cost matrix, and assign costs to imbalanced data features. Finally, use back propagation fine-tuning the parameters of the neural network model to obtain the optimal feature vector. The simulation results show that compared with the FAE algorithm and the NDAE algorithm, the cost-constrained algorithm has a better improvement in intrusion detection accuracy and convergence when faced with multi-dimensional and class imbalanced data.