Resource Efficient Boosting Method for IoT Security Monitoring

Zakariyya, Idris; Al-Kadri, M. Omar; Kalutarage, Harsha

doi:10.1109/ccnc49032.2021.9369620

Cited by 9 publications

(4 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The bijective soft set algorithm [ 45 ] was used to determine the most effective ML model based on accuracy, precision, recall and training time. Zakariyya et al [ 46 ] recommended LGBM as a resource-efficient ML method. Susilo et al [ 47 ] proposed the CNN model for botnet detection in Software-Defined Networks (SDN), and it outperformed the RF model.…”

Section: Review Of Related Workmentioning

confidence: 99%

SMOTE-DRNN: A Deep Learning Algorithm for Botnet Detection in the Internet-of-Things Networks

Popoola

Adebisi

Ande

et al. 2021

Sensors

View full text Add to dashboard Cite

Nowadays, hackers take illegal advantage of distributed resources in a network of computing devices (i.e., botnet) to launch cyberattacks against the Internet of Things (IoT). Recently, diverse Machine Learning (ML) and Deep Learning (DL) methods were proposed to detect botnet attacks in IoT networks. However, highly imbalanced network traffic data in the training set often degrade the classification performance of state-of-the-art ML and DL models, especially in classes with relatively few samples. In this paper, we propose an efficient DL-based botnet attack detection algorithm that can handle highly imbalanced network traffic data. Specifically, Synthetic Minority Oversampling Technique (SMOTE) generates additional minority samples to achieve class balance, while Deep Recurrent Neural Network (DRNN) learns hierarchical feature representations from the balanced network traffic data to perform discriminative classification. We develop DRNN and SMOTE-DRNN models with the Bot-IoT dataset, and the simulation results show that high-class imbalance in the training data adversely affects the precision, recall, F1 score, area under the receiver operating characteristic curve (AUC), geometric mean (GM) and Matthews correlation coefficient (MCC) of the DRNN model. On the other hand, the SMOTE-DRNN model achieved better classification performance with 99.50% precision, 99.75% recall, 99.62% F1 score, 99.87% AUC, 99.74% GM and 99.62% MCC. Additionally, the SMOTE-DRNN model outperformed state-of-the-art ML and DL models.

show abstract

Section: Review Of Related Workmentioning

confidence: 99%

SMOTE-DRNN: A Deep Learning Algorithm for Botnet Detection in the Internet-of-Things Networks

Popoola

Adebisi

Ande

et al. 2021

Sensors

View full text Add to dashboard Cite

show abstract

“…The complete description of this dataset is shown in table 1. [19] 93.13 Zakariyya et al, [10] 89.32 Gomes et al, [20] 95.32 Singh et al, [21] 95.51 Islam et al, [22] 96.19 Singh et al, [23] 90.35 Bhuvaneswari et al, [24] 94.48 Proposed work 96.60…”

Section: ░ 4 Results and Discussionmentioning

confidence: 99%

“…Ensemble learners make it easier to discover abnormal events in IoT network traffic by employing collaborative and optimised forecasting methodologies [8]. Using a decision treebased aggregation model, such as the compact gradient boosting model, improves IoT network detection of anomalies [9], [10] while lowering computing time on resource-limited devices. Furthermore, ensemble algorithms enhance predictive performance and decision-making abilities, similar to human reasoning [11].…”

Section: ░ 2 Literature Reviewmentioning

confidence: 99%

Empowering Smart City IoT Network Intrusion Detection with Advanced Ensemble Learning-based Feature Selection

Merlin,

Ravi

2024

IJEER

View full text Add to dashboard Cite

This study presents an advanced methodology tailored for enhancing the performance of Intrusion Detection Systems (IDS) deployed in Internet of Things (IoT) networks within smart city environments. Through the integration of advanced techniques in data preprocessing, feature selection, and ensemble classification, the proposed approach addresses the unique challenges associated with securing IoT networks in urban settings. Leveraging techniques such as SelectKBest, Recursive Feature Elimination (RFE), and Principal Component Analysis (PCA), combined with the Gradient-Based One Side Sampling (GOSS) technique for model training, the methodology achieves high accuracy, precision, recall, and F1 score across various evaluation scenarios. Evaluation on the UNSW-NB15 dataset demonstrates the effectiveness of the proposed approach, with comparative analysis showcasing its superiority over existing techniques.

show abstract

“…The advantage of the filter-based technique is that it is simple and computationally efficient, but it is unable to exploit the relationship between features, thereby reducing the overall level of accuracy. Examples of filterbased techniques used for feature selection are grid-search [13], Recursive Feature Elimination (RFE) [14], and elastic net [12]. Wrapped-based methods utilize classifier knowledge to determine the optimal feature subset using an evolutionary algorithm to identify optimal solutions by analyzing the search area of a set of solutions (population) [15].…”

Section: Introductionmentioning

confidence: 99%

ANOVA-Artificial Bee Colony Algorithm-Driven Feature Selection for Classifying Downy Mildew Severity in Melon Leaves

Rozikin,

Buono,

Wahjuni

et al. 2023

ISI

View full text Add to dashboard Cite

Diseases of melon plants can cause losses to farmers, such as reduced productivity or even death of melon plants. Downy mildew (DM) is a well-known fast-spreading disease affecting the leaves of melon plants. It is important to determine the level of severity of DM leaf disease so that farmers can take preventive measures according to the severity of DM disease that infects the leaves. Determining the severity of DM disease can be done with experts, but experts have limitations, namely, the availability of experts, and not all areas have leaf disease experts. The stages of this research were data acquisition, preprocessing, feature extraction, feature selection, and classification. Data were taken directly from farmers' gardens and then pre-processed. Color, texture, edge, and entropy features were extracted to obtain the combined feature values. Combined features are prone to redundant and irrelevant features; therefore, feature selection must be performed to obtain the best features. This research proposes an integration concept between the analysis of variance (ANOVA) and artificial bee colony (ABC) optimization, which is named AVABC, and is used as a feature selection algorithm. The test results for the search process time for the eight best features using the AVABC algorithm took 05 minutes 23 s, whereas the test results for the search process time for eight features using ABC with the accuracy model fitness function took 20 h 08 min 55 s. The AVABC feature selection algorithm has the advantage of faster search time for the eight best features.

show abstract

Resource Efficient Boosting Method for IoT Security Monitoring

Cited by 9 publications

References 25 publications

SMOTE-DRNN: A Deep Learning Algorithm for Botnet Detection in the Internet-of-Things Networks

SMOTE-DRNN: A Deep Learning Algorithm for Botnet Detection in the Internet-of-Things Networks

Empowering Smart City IoT Network Intrusion Detection with Advanced Ensemble Learning-based Feature Selection

ANOVA-Artificial Bee Colony Algorithm-Driven Feature Selection for Classifying Downy Mildew Severity in Melon Leaves

Contact Info

Product

Resources

About