optimization methods effectively solve challenging machinelearning problems. optimization is the numerical computation of the system-designed parameter to make predictions on big data. machinelearning algorithms tra...
详细信息
Distributed Denial of Service attack is widely utilized by cyber attackers to target organizations to gain financial advantages. The different organizations aim to tackle these attacks, but manual work or precautions ...
详细信息
With the rapid development of autonomous driving technology, intra-vehicle communication plays a key role in efficient inter-vehicle interaction and internal system synergy. However, the complexity and challenge of in...
详细信息
We present a novel approach in network security using unsupervised online machinelearning method at the edge, through graph learning. The proposed system takes advantage of an online learning paradigm, by collecting ...
详细信息
ISBN:
(纸本)9798400709265
We present a novel approach in network security using unsupervised online machinelearning method at the edge, through graph learning. The proposed system takes advantage of an online learning paradigm, by collecting real network data to build a ground truth of a network's topology, using shallow graph neural networks (GNNs). Our proposed solution includes an edge-based infrastructure, through K3s and Kafka, which could then scale to match the needs of larger networks. We then perform simple cyber-attacks and show how visual analysis can identify malicious behaviors, without any prior labeled data. Our results against simple attacks show promise that improved graph analytics should capture even more complex attack vectors. We then conclude with some suggestions for improved edge deployment, against larger and more complex networks.
Recently, quantum machinelearning is gaining a lot of attention, yet its actual utility in compared to classical machinelearning methods is still unknown. There are signs, meanwhile, that some quantum machine learni...
详细信息
The proceedings contain 253 papers. The topics discussed include: intelligent network traffic control with AI and machinelearning;a dynamic routing switching method based on graph neural network for fire rescue site ...
ISBN:
(纸本)9798331505264
The proceedings contain 253 papers. The topics discussed include: intelligent network traffic control with AI and machinelearning;a dynamic routing switching method based on graph neural network for fire rescue site command battle network;trend of hot data in new media communication based on artificial intelligence;transfer learning for efficient node placement in dynamic wireless sensor networks;design of lightweight decentralized secure communication;improvement of spiking neural network with bit plane coding;optimizing mobile charger scheduling in wireless rechargeable sensor networks using min-heap algorithm;design and application of a virtual simulation system based on wireless sensor networks;deep learning-enhanced adaptive node placement in wireless sensor networks;and weighted deep learning implementation for cognitive radio attack detection.
In recent years, the rapid development of the Internet industry has made the network traffic show a diversified trend. Different application services have different characteristics and have different requirements on n...
详细信息
Wireless Multimedia Sensor networks (WMSN) are autonomous, limited energy and distributed multimedia sensor which can transmit the multimedia data through communicationnetworks. In WMSN, sensor nodes sense data and c...
详细信息
Wireless Multimedia Sensor networks (WMSN) are autonomous, limited energy and distributed multimedia sensor which can transmit the multimedia data through communicationnetworks. In WMSN, sensor nodes sense data and capture multimedia data. The WMSN deployed in many applications such as surveillance, healthcare, military applications and etc., The enhancement of high speed communication in WMSN is increasing demand for multimedia services and applications. The WMSN mostly used for transmission of multimedia services such as audio and video. The volume of multimedia data is increase vastly as well as utilization of bandwidth and energy also increase. The limited energy WMSN exhaust rapidly and its leads to network failure. Previous extensive research carried out to maximize throughput by the optimal utilization of network bandwidth. Also improved QoS by the novel routing algorithm in WMSN. To address the limitation of existing frameworks, in this research paper proposed an emerging energy management framework to reduce energy utilization and improve network life time. The proposed E2M framework implementation is made using NS2. The empirical results shown that proposed framework outperforms with comparison of existing frameworks. The performance evaluated in terms of PDR, Delay, Throughput and Energy. The proposed E2M Framework reduce 5.74 ms of network delay, improved 5.3% of PDR, Throughput improved to 77KB and reduce energy consumption of 0.39 kJ.
Although traditional protection methods can improve the security of the network to a certain extent, its effect is often limited in the face of complex network environment and various attacks. Therefore, it is an impo...
详细信息
This paper proposes an understandable neural network whose score function is modeled as an additive sum of univariate spline functions. It extends usual understandable models like generative additive models, spline-ba...
详细信息
This paper proposes an understandable neural network whose score function is modeled as an additive sum of univariate spline functions. It extends usual understandable models like generative additive models, spline-based models, and neural additive models. It is shown that this neural network can be approximated by a logistic regression whose inputs are obtained with a non-linear preprocessing of input data. This preprocessing depends on the neural network initialization but this paper establishes that it can be replaced by a non random kernel-based preprocessing that no longer depends on the initialization. Hence, the convergence of the training process is guaranteed and the solution is unique for a given training dataset.
暂无评论