A network intrusion detection model that fuses a convolutional neural and gated recurrent unit is proposed to address the problems associated with low accuracy of existing models for multiple classification intrusions class imbalance data detection. In this model, hybrid sampling algorithm combining Adaptive Synthetic Sampling (ADASYN) Repeated Edited nearest neighbors (RENN) used sample processing solve problem positive negative in original dataset. The feature selection carried out by...

At present, there are many techniques for cyber security defense such as firewall, intrusion detection and cryptography. Despite decades of studies experiences on this issue, still exists a problem that we always pay great attention to technology while overlooking strategy. In the traditional warfare, level decision-making formulation optimal strategies have effect warfare result. Similarly, timeliness quality in attack-defense also make significance. Since attackers defenders oppositional,...

As a security defense technique to protect networks from attacks, network intrusion detection model plays crucial role in the of computer systems and networks. Aiming at shortcomings complex feature extraction process insufficient information existing models, an named FCNN-SE, which uses fusion convolutional neural (FCNN) for stacked ensemble (SE) classification, is proposed this paper. The mainly includes two parts, classification. Multi-dimensional features traffic data are first extracted...

JavaScript has been widely used on the Internet because of its powerful features, and almost all websites use it to provide dynamic functions. However, these natures also carry potential risks. The authors malicious scripts started using launch various attacks, such as Cross-Site Scripting (XSS), Cross-site Request Forgery (CSRF), drive-by download attack. Traditional script detection relies expert knowledge, but even for experts, this is an error-prone task. To solve problem, many...

Aiming at the inherent problems of swarm intelligence algorithm, such as falling into local extremum in early stage and low precision later stage, this paper proposes an improved sparrow search algorithm (ISSA). Firstly, we introduce idea flight behavior bird SSA to keep diversity population reduce probability optimum; Secondly, creatively crossover mutation genetic get better next-generation population. These two improvements not only all times but also make up for defect that is easy fall...

Abstract Fennel contains many antioxidant and antibacterial substances, it has very important applications in food flavoring other fields. The kinds contents of chemical substances fennel vary from region to region, which can affect the taste efficacy its derivatives. Therefore, is great significance accurately classify origin fennel. Recently, detection methods based on deep networks have shown promising results. However, existing spend a relatively large time cost, drawback that fatal for...

Previous studies have attempted to find autonomic differences of the cardiac system between congestive heart failure (CHF) disease and healthy groups using a variety algorithms pattern recognition. By comparing previous literature, we found that there are two shortcomings: (1) focused on improving accuracy models, but number features used has mostly exceeded 10, leading poor generalization performance; (2) works rarely distinguish severity levels CHF disease. In order make up for these...

With the rapid development of e-commerce systems, centralized service model gradually fails to meet needs SMEs. In existing system, users’ transaction data and reputation scores are stored in a cloud server, which has high storage cost, low processing efficiency, is vulnerable attacks leaks. However, decentralized systems more than its system store average credit score evaluation receiving unfair evaluations against risk. The malicious nodes no disciplinary measures, not conducive...

Abstract There is a class-imbalance problem that the number of minority class samples significantly lower than majority in common network traffic datasets. Class-imbalance phenomenon will affect performance classifier and reduce robustness to detect unknown anomaly detection. And distribution continuous features dataset does not follow Gaussian distribution, which bring great difficulties intrusion We propose Conditional Wasserstein Variational Autoencoders with Generative Adversarial...

Imbalanced datasets greatly affect the analysis capability of intrusion detection models, biasing their classification results toward normal behavior and leading to high false-positive false-negative rates. To alleviate impact class imbalance on accuracy network models improve effectiveness, this paper proposes a method based feature selection-conditional Wasserstein generative adversarial (FCWGAN) bidirectional long short-term memory (BiLSTM). The uses XGBoost algorithm with Spearman's...

Network intrusion detection system plays an important role in network security. Aiming at the problem that it is difficult to extract subtle features process of detection, a model based on neural feature extraction and particle swarm optimization algorithm optimize support vector machine proposed. In this method, one-dimensional data constructed into two-dimensional matrix data, which used as input convolutional network, information extracted from full connection layer. Finally, accuracy...

Support Vector Machine (SVM) is a relatively novel classification technology, which has shown higher performance than traditional learning methods in many applications. Therefore, some security researchers have proposed an intrusion detection method based on SVM. However, the SVM algorithm very sensitive to choice of kernel function and parameter adjustment. Once selection unscientific, it will lead poor accuracy. To solve this problem, paper presents Grey Wolf Optimizer Algorithm Particle...

To resolve the problems of low prediction accuracy and slow convergence speed traditional extreme learning machines in network security situation methods, we combine a meta-heuristic search algorithm with neural networks propose method based on improved sparrow optimization an machine. Firstly, initial population is initialized by cat-mapping chaotic sequences to enhance randomness ergodicity improve global ability algorithm. Secondly, Cauchy mutation tent chaos disturbance are introduced...