ResInceptNet-SA: A Network Traffic Intrusion Detection Model Fusing Feature Selection and Balanced Datasets

Applied Sciences, Journal Year: 2025, Volume and Issue: 15(2), P. 956 - 956

Published: Jan. 19, 2025

Network intrusion detection models are vital techniques for ensuring cybersecurity. However, existing face several challenges, such as insufficient feature extraction capabilities, dataset imbalance, and suboptimal accuracy. In this paper, a new type of model (ResIncepNet-SA) based on InceptionNet, Resnet, convolutional neural networks with self-attention mechanism was proposed to detect network intrusions. The used the PCA-ADASYN algorithm compress traffic features, extract high-correlation datasets, oversample balance datasets classify abnormal traffic. experimental results show that accuracy, precision, recall, F1-score ResIncepNet-SA using NSL-KDD reach 0.99366, 0.99343, 0.99339, 0.99338, respectively. This enhances accuracy outperforms when applied imbalanced offering solution detection.

Language: Английский

---

Artificial Intelligence in Cybersecurity: A Comprehensive Review and Future Direction

Applied Artificial Intelligence, Journal Year: 2024, Volume and Issue: 38(1)

Published: Dec. 10, 2024

As cybercrimes are becoming increasingly complex, it is imperative for cybersecurity measures to become more robust and sophisticated. The crux lies in extracting patterns or insights from data build data-driven models, thus making the security systems automated intelligent. To comprehend analyze data, several Artificial Intelligence (AI) methods such as Machine Learning (ML) techniques, employed monitor network environments actively combat cyber threats. This study explored various AI techniques how they applied cybersecurity. A comprehensive literature review was conducted, including a bibliometric analysis systematic following PRISMA (Preferred Reporting Items Systematic Reviews Meta-Analyses) guidelines. Using extracted two main scholarly databases: Clarivate's Web of Science (WoS) Scopus, this article examines relevant academic understand diverse ways which strengthen measures. These applications range anomaly detection threat identification predictive analytics incident response. total 14,509 peer-reviewed research papers were identified 9611 Scopus database 4898 WoS database. further filtered, 939 eventually selected used. offers into effectiveness, challenges, emerging trends utilizing purposes.

Language: Английский

---

A comparative assessment of machine learning algorithms in the IoT-based network intrusion detection systems

Decision Analytics Journal, Journal Year: 2024, Volume and Issue: 11, P. 100478 - 100478

Published: May 15, 2024

The rapid increase in online risks is a reflection of the exponential growth Internet Things (IoT) networks. Researchers have proposed numerous intrusion detection techniques to mitigate harm caused by these threats. Enterprises use systems (IDSs) and prevention (IPSs) keep their networks safe, stable, accessible. Network solutions lately integrated powerful Machine Learning (ML) safeguard IoT Selecting proper data features for effectively training such ML models critical maximizing accuracy computational efficiency. However, efficiency degrades high-dimensional spaces, it crucial suitable feature extraction method eliminate extraneous from classification procedure. false positive rate many ML-based IDSs also rise when samples used train are unbalanced. This study provides detailed overview UNSW-NB15(DS-1) NF-UNSWNB15(DS-2) datasets detection, which will be utilized develop evaluate our models. In addition, this model uses MaxAbsScaler algorithm implement filter-based scaling strategy . Then, condensed set perform several techniques, including Support Vector Machines (SVM), K-nearest neighbors (KNN), Logistic Regression (LR), Naive Bayes (NB), Decision Tree (DT), Random Forest (RF), considering multiclass classification. Accuracy tests scheme were improved 60% 94% using MaxAbsScaler-based method.

Language: Английский

---

Machine Learning Insights: Exploring Key Factors Influencing Sale-to-List Ratio—Insights from SVM Classification and Recursive Feature Selection in the US Real Estate Market

Buildings, Journal Year: 2024, Volume and Issue: 14(5), P. 1471 - 1471

Published: May 18, 2024

The US real estate market is a complex ecosystem influenced by multiple factors, making it critical for stakeholders to understand its dynamics. This study uses Zillow Econ (monthly) data from January 2018 October 2023 across 100 major regions gathered through Metropolitan Statistical Area (MSA) and advanced machine learning techniques, including radial kernel Support Vector Machines (SVMs), used predict the sale-to-list ratio, key metric that indicates health competitiveness of estate. Recursive Feature Elimination (RFE) identify influential variables provide insight into Results show SVM achieves approximately 85% accuracy, with temporal indicators such as Days Pending Close, pricing dynamics Listing Price Cut Share Listings Cut, rental conditions captured Observed Rent Index (ZORI) emerging factors influencing ratio. comparison between alphas RFE highlights importance time, price, in understanding trends. underscores interplay these provides actionable insights stakeholders. By contextualizing findings within existing literature, this emphasizes considering housing analysis. Recommendations include using inform strategies negotiation tactics. adds body knowledge research foundation informed decision-making ever-evolving landscape.

Language: Английский

---

An AI-Driven Based Cybersecurity System for Network Intrusion Detection System in Hybrid with EPO and CNNet-LAM

Published: Feb. 9, 2024

The proliferation of large data made possible by ubiquitous internet use has led to an uptick in cyberattacks, despite the AI-based security keys like intrusion detection systems (IDS). Improved classification is one benefit suggested system's foundation deep learning (DL) and Convolutional Neural Networks (CNNs). With IDSAI dataset, this research takes a close look at systems. Z-Score Normalisation Min-Max are used for preparation. Picking out most relevant characteristics from preprocessed next step after preprocessing. As result, feature selection method makes optimisation known as Eagle Perching Optimisation (EPO) Algorithm. with Long Short-Term Memory Attention Mechanism (CNNet-LAM) selection. It common practice employ EPO during hyperparameter tweaking due its efficacy. Classification issues may be effectively resolved using CNNet-LAM hybrid model. model consistently surpasses competition, according testing data, it can predict varying time delays accuracy 99.31%.

Language: Английский

---

Strengthening Cybersecurity using a Hybrid Classification Model with SCO Optimization for Enhanced Network Intrusion Detection System

Published: March 15, 2024

Polymorphic malware and encrypted traffic hinder Network Intrusion Detection Systems (NIDS) from detecting complex attacks. Cybercriminals exploit NIDS algorithm vulnerabilities, showing how attack tactics cybersecurity defenses change. This study suggests improving Systems. A thorough preprocessing phase with normalization functions improves data accuracy consistency. The Single Candidate Optimization (SCO) feature selection optimizes efficacy. hybrid model using Wavelet Transform, Long Short-Term Memory (LSTM), Artificial Neural Networks (ANN) is used for classification because it can identify sequential dependencies in network data. second SCO iteration hyperparameter tuning performance refines. evaluation stage uses the BoT-IoT dataset, a prominent benchmark. improve optimization to create more accurate cyberattack-resistant NIDS. method's 99.6% confirmed by experiments evaluations. shows effective compared current models, which strengthens against changing landscapes. Similar trends are seen F1-scores, range 96.3% (ResNet50) 99.4% (Proposed model). Projected performs exceptionally well terms of stands out highest values all metrics.

Language: Английский

---

An improved federated transfer learning model for intrusion detection in edge computing empowered wireless sensor networks

Concurrency and Computation Practice and Experience, Journal Year: 2024, Volume and Issue: 36(24)

Published: Aug. 15, 2024

Summary Intrusion Detection (ID) is a critical component in cybersecurity, tasked with identifying and thwarting unauthorized access or malicious activities within networked systems. The advent of Edge Computing (EC) has introduced paradigm shift, empowering Wireless Sensor Networks (WSNs) decentralized processing capabilities. However, this transition presents new challenges for ID due to the dynamic resource‐constrained nature environments. In response these challenges, study pioneering approach: an Improved Federated Transfer Learning Model. This model integrates pre‐trained ResNet‐18 transfer learning meticulously designed Convolutional Neural Network (CNN), tailored intricacies NSL‐KDD dataset. collaborative synergy models culminates System (IDS) impressive accuracy 96.54%. Implemented Python, proposed not only demonstrates its technical prowess but also underscores practical applicability fortifying EC‐empowered WSNs against evolving security threats. research contributes ongoing discourse on enhancing cybersecurity measures emerging computing paradigms.

Language: Английский

---

Improving Network Security with Gradient Boosting from KDD Cup Dataset

SN Computer Science, Journal Year: 2024, Volume and Issue: 5(7)

Published: Sept. 13, 2024

Language: Английский

---

Cybersecurity Using Hybrid Type Model for Classification Through SCO Optimization Technique

Published: Feb. 9, 2024

The effectiveness of Network Intrusion Detection Systems (NIDS) in recognizing complex attacks is hindered by evasion strategies like polymorphic malware and encrypted traffic. Vulnerabilities NIDS algorithms are regularly exploited cybercriminals, highlighting the dynamic relationship between changing attack tactics cybersecurity defenses. An enhanced approach to strengthen suggested this study. methodology begins with a rigorous preprocessing phase that includes normalization functions progress accuracy consistency input data. Optimizing NIDS's efficacy aim Single Candidate Optimization (SCO) algorithm for feature selection. We utilize hybrid model incorporates ANN, Wavelet Transform, Long Short-Term Memory (LSTM) classification phase. This designed identify sequential dependencies network traffic data we have. A second iteration SCO devoted hyperparameter optimization order attain optimal performance further optimize model. For our review, opted use BoT-IoT dataset because it gold standard field. demonstrates how can improve selection, leading more accurate cyberattack-resistant NIDS. Consequences from experiments evaluations demonstrate proposed strategy effective; achieves remarkable 99.6 percent accuracy. proves successful judgement current representations, which significant step towards consolidation defenses against threat landscapes.

Language: Английский

---

Enhancing iot network security through advanced data preprocessing and hybrid firefly-salp swarm optimized deep cnn-based intrusion detection

ITEGAM- Journal of Engineering and Technology for Industrial Applications (ITEGAM-JETIA), Journal Year: 2024, Volume and Issue: 10(47)

Published: Jan. 1, 2024

This concept addresses the imperative need for robust Intrusion Detection system (IDs) in Internet of Things (IoT) networks by presenting a comprehensive approach that integrates advanced data preprocessing techniques and Deep Convolutional Neural Network (DCNN) based IDS. The process commences with raw inherently noisy generated IoT sensors. To fortify detection capabilities, sequence steps is applied, including cleaning, one-hot encoding normalization, ensuring prepared resilient to outliers irrelevant information while being conducive Learning (DL) models. core proposed DCNN, adept at capturing sequential patterns within diverse dynamic data. further optimize performance hybrid firefly-salp swarm optimization algorithm employed. leverages strengths both Firefly salp (FFA-SSA), enhancing model's ability identify potential security threats effectively. synergy nature-inspired methods not only strengthens posture but also contributes resilience adaptability intrusion systems. presented signifies crucial step towards more secure deployments, acknowledging pivotal role played innovative preparing optimizing deep learning models enhanced cybersecurity.

Language: Английский

---