Dimensionality reduction of the attributes using fuzzy optimized independent component analysis for a Big Data Intrusion Detection System
Annotation
Big data cybersecurity has garnered more attraction in recent years with the development of advanced machine learning and deep learning classifiers. These new classifier algorithms have significantly improved Intrusion Detection Systems (IDS). In these classifiers, the performance is positively influenced by high relevant features while less relevant features negatively influence the performance. However, considering all the attributes, especially the high dimensional attributes, increases computational complications. Hence it is essential to diminish the dimensionality of the attributes to improve the classifier performance. To achieve this objective, an efficient dimensionality reduction approach is presented through the development of the Fuzzy Optimized Independent Component Analysis (FOICA) technique. The standard Independent Component Analysis (ICA) is coupled with the fuzzy entropy to transform the high dimension attributes into low dimension attributes and helps in selecting high informative low-dimensional attributes. These selected features are fed to efficient hybrid classifiers namely Hyper-heuristic Support Vector Machines (HH-SVM), Hyper-Heuristic Improved Particle Swarm Optimization based Support Vector Machines (HHIPSO-SVM) and Hyper-Heuristic Firefly Algorithm based Convolutional Neural Networks (HHFA-CNN) to classify the cybersecurity data to identify the intrusions. Experiments are conducted over two cybersecurity datasets and real-time laboratory data whose outcomes specify the supremacy of the suggested IDS model based on FOICA dimensionality reduction.
Keywords
Постоянный URL
Articles in current issue
- A study of a silicone film deposited on quartz glass under laser radiation
- Optical composites based on organic polymers and semiconductor pigments
- A new algorithm for the identification of sinusoidal signal frequency with constant parameters
- A study of silicon p-n structures with mono and multifacial photosensitive surfaces
- Detection of yawning in driver behavior based a convolutional neural network
- A Game Theory approach for communication security and safety assurance in cyber-physical systems with Reputation and Trust-based mechanisms
- A study of the influence of human factors on the speed of urban rail transport
- An algorithm for detecting RFID-duplicates
- Reduction of LSB detectors set with definite reliability
- Classification of objects in images with distortions based on a two-stage topological analysis
- An optimal swift key generation and distribution for QKD
- A study of vectorization methods for unstructured text documents in natural language according to their influence on the quality of work of various classifiers
- Recognition the emotional state based on a convolutional neural network
- Intellectualization of personnel development management in high-tech service-oriented companies
- A study of the efficiency of the magnetic compass correction system
- A new analytical model of drain current and small signal parameters for AlGaN-GaN high-electron-mobility transistors
- Imputation and system modeling of acid-base state parameters for different groups of patients
- Construction of movement trajectories for objects based on the Dubins car problem, taking into account constant external influences
- A mathematical model of an epidemic with an arbitrary law of recovery
- Simulation of the pulsed outflow of air and fine powder mixture, partially filling the discharge channel
- Vectorized numerical algorithms for the solution of continuum mechanics problems
- A comparative analysis of computational intelligence algorithms for estimation of LTE channels
- Implementation of a clinical decision support system to improve the medical data quality for hypertensive patients