A comparative study of Machine learning algorithms for VANET networks

Ftaimi, Sara; Mazri, Tomader

doi:10.1145/3386723.3387829

Cited by 21 publications

(13 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We will compare the SVM results with the neural network results. we chose to compare the neural network with SVM because, based on the comparative study conducted in [3], the neural network also gives good results in handling outliers, and also it is a very fast algorithm and can also handle multidimensional data and does not need much memory.…”

Section: Resultsmentioning

confidence: 99%

“…The high mobility of the nodes is one of the most relevant features of the VANET network vehicles are moving continuously in many directions and at varying speeds. The topology of the vanet network changes rapidly due to the fact that the vehicle can join or leave a cluster of vehicles in a very short time [3].Owing to the two abovementioned features, combined with climate change and traffic congestion, vehicles can frequently become disconnected from the network. VANET has no energy or storage problems [4][5][6], unlike MANET networks, where energy constraint is a challenge.…”

Section: Vanet Networkmentioning

confidence: 99%

“…The SVM algorithm can be used to classify linear and non-linear data. Therefore, it can carry out classification in a non-linear way using the kernel trick [3,17].The SVM can be used to detect network intrusion [18] and spoofing attacks [19]. Steven So, Prinkle Sharma, and Jonathan Petit [20] have used security messages to identify classes and types of attacks using SVM and KNN algorithms.…”

Section: Support Vector Machine (Svm)mentioning

confidence: 99%

“…However, the performance of SVM is much faster than FFNN. The choice of SVM algorithms was based on the comparative study conducted in [3] found that SVM is an algorithm that can handle multidimensional data and does not need much memory, and can detect outliers because the outlier detection is very important since the outliers can make false the classification model. Also, SVM can be used for both the separable and non-linearly separable data and work very well with a higher dimension.…”

Section: Support Vector Machine (Svm)mentioning

confidence: 99%

See 3 more Smart Citations

Handling Priority Data in Smart Transportation System by using Support Vector Machine Algorithm

Ftaimi¹,

Mazri²

2020

Adv. sci. technol. eng. syst. j.

Self Cite

View full text Add to dashboard Cite

Section: Resultsmentioning

confidence: 99%

Section: Vanet Networkmentioning

confidence: 99%

Section: Support Vector Machine (Svm)mentioning

confidence: 99%

Section: Support Vector Machine (Svm)mentioning

confidence: 99%

See 2 more Smart Citations

Handling Priority Data in Smart Transportation System by using Support Vector Machine Algorithm

Ftaimi¹,

Mazri²

2020

Adv. sci. technol. eng. syst. j.

Self Cite

View full text Add to dashboard Cite

“…Deep Learning, a sub field of machine learning helps in more precise traffic flow prediction (Wang et al 2019). Also, deep learning approaches hold promising outcomes to solve complex computation and big data analysis in intelligent transportation applications (Ftaimi and Mazri 2020;Liu and Shoji 2019). The traffic control models are discriminated as short term and long-term prediction.…”

Section: Related Workmentioning

confidence: 99%

QOS aware and secured routing algorithm using machine intelligence in next generation VANET

Nadarajan

Kaliyaperumal

2021

Int J Syst Assur Eng Manag

View full text Add to dashboard Cite

Vehicular ad-hoc networks have been emerging as an inevitable research topic in recent years due to its applications in Intelligent Transportation System. Establishing end-to-end communication and high authenticate information transfer in vehicular networks remains to be a real challenge due to the dynamic topology change, vehicle interspacing, high vehicle density and non-intelligent authenticated routing mechanisms. To address these aforementioned, a unique stochastic based adaptive routing integrated with secured 3D logistic chaotic maps and traffic flow predictions has been developed and referred in this paper as Stochastic Chaos based Adaptive Routing with Prediction (SCARP). The SCARP can perform the good traffic flow predictions using strong deep learning networks which recommends a stochastic based node discovery routing principle. This approach has reduced connectivity loss, minimized delay and as well provides chaotic authentication to form an efficient and safe data transmission among the vehicular nodes. The performance testing of this proposed method is done with a few months of traffic data obtained from the layout Puducherry U.T, India. These experimentations are carried out in SUMO-OMNET?? environment and compared with existing learning and routing algorithms. Simulation results demonstrates that the proposed algorithm has outperformed the other existing algorithms in terms of prediction accuracy, sensitivity, average delay, packet delivery ratio. Also the transmission performance of the proposed algorithm has more robustness with the high intensity DoS attacks.

show abstract