Leveraging stacking machine learning models and optimization for improved cyberattack detection DOI Creative Commons

Neha Pramanick,

Jimson Mathew, Shitharth Selvarajan

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: May 14, 2025

The ever-growing number of complex cyber attacks requires the need for high-level intrusion detection systems (IDS). While available research deals with traditional, hybrid, and ensemble methods network data analysis, serious challenges are still being met in terms producing robust highly accurate systems. There high hurdles managing high-dimensional traffic since current methodologies limited dealing imbalanced issues minority classes versus majority false positive rate classification accuracy. This study introduces an innovative framework that directly addresses these persistent through a novel approach to detection. proposed method integrates two ML models: J48 ExtraTreeClassifier classification. Besides, we propose improved equilibrium optimizer (EO) whereby previous EO is modified. In this enhanced (EEO), Fisher score accuracy K-Nearest Neighbors (KNN) algorithm select attributes optimally, whereas synthetic oversampling technique combined iterative partitioning filters (SMOTE-IPF) used provide class balancing. KNN also imputation improve overall system superior performance has been validated experimentally on several benchmark datasets, i.e., NSL-KDD, UNSW-NB15, achieving 99.7% 98.1% F1 99.6 98.0 respectively. By subjecting comparative analysis recent state-of-the-art works, paper shown methodology yields better improvement feature selection precision accuracy, handling instance, less demanding storage computational efficiency.

Language: Английский

Privacy-preserving approach for IoT networks using statistical learning with optimization algorithm on high-dimensional big data environment DOI Creative Commons
Fatma S. Alrayes, Mohammed Maray, Asma Alshuhail

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: Jan. 27, 2025

Language: Английский

Citations

2
A Pervasive Model to Evaluate and Ranking the Best 5G Mobile by Cross‐Entropy and TOPSIS Method DOI Open Access
Pankaj Prasad Dwivedi, Dilip Kumar Sharma

International Journal of Communication Systems, Journal Year: 2025, Volume and Issue: 38(4)

Published: Jan. 16, 2025

ABSTRACT The development of the new 5G mobile generation is among most significant and recent breakthroughs in communications information technology. Compared to older networks, wireless cellular technology offers more capacity, dependable connections, quicker upload download rates. newest phones, or have significantly speeds are dependable. They potential completely change way we access social media, websites, information. This study was carried out address issue that many users frequently face when deciding which phone best for them based on factors. contributes putting a strategy assessing ranking various possibilities light consumers' order predilections multicriteria decision‐making (MCDM). In this study, studied 15 different mobiles data from companies with 10 criteria. First, appealing criteria affect person's decision about determined. accomplished by survey target audience, expertise industry professionals area, research literature. For alternatives, TOPSIS approach employed, relative weights assessment calculated cross‐entropy approach, terms MCDM. demonstrates usefulness suggested strategy.

Language: Английский

Citations

1
Smart Farming Revolution: A Cutting-Edge Review of Deep Learning and IoT Innovations in Agriculture DOI

J. Siva Prashanth,

G. Bala Krishna,

A. V. Krishna Prasad

et al.

Operations Research Forum, Journal Year: 2025, Volume and Issue: 6(1)

Published: March 14, 2025

Language: Английский

Citations

1
Enhanced novelty approaches for resource allocation model for multi-cloud environment in vehicular Ad-Hoc networks DOI Creative Commons

R. Augustian Isaac,

P. Sundaravadivel,

V. Marx

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: March 19, 2025

As the number of service requests for applications continues increasing due to various conditions, limitations on resources provide a barrier in providing with appropriate Quality Service (QoS) assurances. result, an efficient scheduling mechanism is required determine order handling application requests, as well use broadcast media and data transfer. In this paper innovative approach, incorporating Crossover Mutation (CM)-centered Marine Predator Algorithm (MPA) introduced effective resource allocation. This strategic allocation optimally schedules within Vehicular Edge computing (VEC) network, ensuring most utilization. The proposed method begins by meticulous feature extraction from network model, attributes such mobility patterns, transmission medium, bandwidth, storage capacity, packet delivery ratio. For further analysis Elephant Herding Lion Optimizer (EHLO) algorithm employed pinpoint critical attributes. Subsequently Modified Fuzzy C-Means (MFCM) used vehicle clustering centred selected These clustered characteristics are then transferred stored cloud server infrastructure. performance methodology evaluated using MATLAB software simulation method. study offers comprehensive solution challenge Cloud Networks, addresses burgeoning demands modern while QoS assurances signifies significant advancement field VEC.

Language: Английский

Citations

0
Enhancing the e-commerce shopping experience with IoT-enabled smart carts in smart stores DOI Creative Commons
Jian Wang

Discover Internet of Things, Journal Year: 2025, Volume and Issue: 5(1)

Published: April 1, 2025

Language: Английский

Citations

0
An Efficient Multi‐Core DSP Power Management Controller DOI Creative Commons

Jian Huang,

Huili Wang, G. H. Gong

et al.

Engineering Reports, Journal Year: 2025, Volume and Issue: 7(4)

Published: April 1, 2025

ABSTRACT Today's society has entered a digital era, and the use of DSP is becoming increasingly frequent important. In order to achieve market targets high energy efficiency, it necessary integrate low‐power design from chip stage. Based on FT‐xDSP architecture, this work designs power management controller for suitable multi‐core multi‐integrated peripherals perspective control in This can precisely supply, clock, memory each module introduces clamp unit solve problem possible glitches during asynchronous reset ensures that system no overflow redundant requests. Additionally, configurable state transition counter also set up avoid insufficient time low‐speed or long waiting high‐speed peripherals. After pre‐tapeout experiment post‐tapeout testing data analysis, above new manager excellent performance. low consumption chip, overall core reduced by over 95%, which great significance achieving high‐efficiency processor chips.

Language: Английский

Citations

0
An efficient electricity theft detection based on deep learning DOI Creative Commons
Nada M. Elshennawy, Dina M. Ibrahim,

Ahmed M. Gab Allah

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: April 15, 2025

Abstract Electrical theft is a pervasive issue that has detrimental impacts on both utility companies and electrical consumers worldwide. It undermines the economic growth of businesses, poses risks, affects customers’ expensive energy bills. Smart grids produce vast quantities data, including consumer usage data which crucial for identifying instances theft. Machine learning deep algorithms may use this to identify This research presents new approach using convolutional neural networks long-short-term memory extract abstract characteristics from power consumption improve accuracy detection electricity users. We mitigate dataset shortcomings, such as incomplete imbalanced class distribution, by LoRAS augmentation. The method’s efficiency evaluated authentic obtained State Grid Corporation China. Finally, we demonstrate competitiveness our when compared other approaches have been assessed same dataset. During validation process, attained 97% rate, surpassing highest reported in previous studies 1%. values 98.75%, 95.45%, 97.7%, along with corresponding recall F1 scores. findings indicate suggested surpasses existing state-of-arts methods.

Language: Английский

Citations

0
Leveraging stacking machine learning models and optimization for improved cyberattack detection DOI Creative Commons

Neha Pramanick,

Jimson Mathew, Shitharth Selvarajan

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: May 14, 2025

The ever-growing number of complex cyber attacks requires the need for high-level intrusion detection systems (IDS). While available research deals with traditional, hybrid, and ensemble methods network data analysis, serious challenges are still being met in terms producing robust highly accurate systems. There high hurdles managing high-dimensional traffic since current methodologies limited dealing imbalanced issues minority classes versus majority false positive rate classification accuracy. This study introduces an innovative framework that directly addresses these persistent through a novel approach to detection. proposed method integrates two ML models: J48 ExtraTreeClassifier classification. Besides, we propose improved equilibrium optimizer (EO) whereby previous EO is modified. In this enhanced (EEO), Fisher score accuracy K-Nearest Neighbors (KNN) algorithm select attributes optimally, whereas synthetic oversampling technique combined iterative partitioning filters (SMOTE-IPF) used provide class balancing. KNN also imputation improve overall system superior performance has been validated experimentally on several benchmark datasets, i.e., NSL-KDD, UNSW-NB15, achieving 99.7% 98.1% F1 99.6 98.0 respectively. By subjecting comparative analysis recent state-of-the-art works, paper shown methodology yields better improvement feature selection precision accuracy, handling instance, less demanding storage computational efficiency.

Language: Английский

Citations

0