With the advancement of mobile communication technology and the continuous development of network infrastructure, people increasingly enjoy the convenience brought by mobile communication and the internet. The user...
详细信息
ISBN:
(纸本)9798400710353
With the advancement of mobile communication technology and the continuous development of network infrastructure, people increasingly enjoy the convenience brought by mobile communication and the internet. The user's network experience is also increasingly valued by mobile operators. To help further enhance the quality of network services and market operations, the current study focuses on customer satisfaction as an important indicator. Firstly, feature engineering is used to better anchor the key factors and improve the performance of the model. Secondly, various models such as Tree-based models and Support Vector machine (SVM) model were used for data processing, and Bayesian parameter tuning methods for model optimization. Finally, a predictive model is established using the Stacking ensemble method to integrate the models. It turns out that the ensemble model has a better performance to predict customers' satisfaction than any single model.
We image the structure of the complex chiral molecule Fenchone (C10H16O) from laser-induced electron diffraction data by applying a machinelearning algorithm with a convolutional neural network (CNN).
We image the structure of the complex chiral molecule Fenchone (C10H16O) from laser-induced electron diffraction data by applying a machinelearning algorithm with a convolutional neural network (CNN).
One of the most used techniques to improve a machinelearning model is to gather more data. An interesting field in machinelearning is Sequence Modelling, having Natural Language Processing as the peak of the field. ...
详细信息
One of the most used techniques to improve a machinelearning model is to gather more data. An interesting field in machinelearning is Sequence Modelling, having Natural Language Processing as the peak of the field. The capabilities of Quantum Computing have been growing recently entering the novel field of Quantum machinelearning. In this paper, we propose a Quantum Natural Language Processing classification model named Strongly Entangling Neural network. This model leverages the quantum advantage to imitate part of the behavior of a Recurrent Neural network to process text data into the circuit and perform the classification task. This is accomplished by representing our data in a quantum circuit that relies heavily on the entanglement property of qubits. The results of our model have very favorable metrics, particularly obtaining a 97.70% of accuracy.
A network of commonplace physical things with electronic parts, such as sensors, processors, and wireless communication modules is referred to as the "Internet of Things" (IoT). They can communicate with one...
详细信息
A network of commonplace physical things with electronic parts, such as sensors, processors, and wireless communication modules is referred to as the "Internet of Things" (IoT). They can communicate with one another and various other systems and devices via these components and the internet. It's important to note that the IoT lacks internal security. No organization is in charge of handling the enormous amount of data created by IoT devices since distributed ledger technology, which is a component of blockchain, is decentralized. Decentralization improves privacy and security. By making it extremely difficult for unauthorized users to alter current data record. Blockchain technology is used to safely store IoT data. Secure machine-to-machine transactions are made possible by the IoT and blockchain, boosting security and transparency for all parties involved. IoT devices can interact with private blockchain networks to guarantee the accuracy of transaction records by utilizing the features such as immutability and data encryption. This paper examines the business strategy that can aid in preventing the manipulation of sensitive data.
The proceedings contain 27 papers. The topics discussed include: a load prediction method for data transmission in optical communicationnetworks based on LSTM model;Bayesian-LightGBM classification prediction model o...
ISBN:
(纸本)9798400711688
The proceedings contain 27 papers. The topics discussed include: a load prediction method for data transmission in optical communicationnetworks based on LSTM model;Bayesian-LightGBM classification prediction model of tennis match situation change;supplement to decomposition of mean squared error of weighted geometric mean combined forecasting method;reconstructing cosmic observations with artificial neural networks: a test of Hubble constant;a routing method for flying ad hoc networks based on node degree estimation and game-theoretic forwarding strategy;the study of a distributed networking method based on an improved raft protocol;research on network security spatial data asset protection technology based on deep learning;and introduction to integration technology based on dispatch telephone and human machine workstation.
Modern data centers are witnessing fast-growing east-west traffic on their network infrastructure due to the highly distributed data center applications. Motivated by the heterogeneity of such application workloads, w...
详细信息
Modern data centers are witnessing fast-growing east-west traffic on their network infrastructure due to the highly distributed data center applications. Motivated by the heterogeneity of such application workloads, we propose in this article an extensible network management architecture called MAGNet which enables application-aware intra-data center networking. The crux of MAGNet is the smart endpoint residing within end-hosts, which is empowered by machinelearning combined with lightweight workload tracing to detect workload identities and enable workload-dependent packet tagging. The centralized management plane interface of MAGNet allows network functions to interpret packet tags and perform application-aware packet processing. We demonstrate the feasibility of the architecture via prototype implementation and extensive use case evaluation. Our experiments show that the smart endpoint can fingerprint many real-world applications with 99 percent accuracy only at 1-2 percent additional CPU, and that application-aware data plane can potentially bring substantial benefits in terms of security (e.g., via identity-based microsegmentation), CPU usage (e.g., for intrusion detection) and network latency (e.g., via TCP stack customization).
A prominent paradigm for graph neural networks is based on the message-passing framework. In this framework, information communication is realized only between neighboring nodes. The challenge of approaches that use t...
详细信息
A prominent paradigm for graph neural networks is based on the message-passing framework. In this framework, information communication is realized only between neighboring nodes. The challenge of approaches that use this paradigm is to ensure efficient and accurate long-distance communication between nodes, as deep convolutional networks are prone to oversmoothing. In this paper, we present a novel method based on time derivative graph diffusion (TIDE) to overcome these structural limitations of the message-passing framework. Our approach allows for optimizing the spatial extent of diffusion across various tasks and network channels, thus enabling medium and long-distance communication efficiently. Furthermore, we show that our architecture design also enables local message-passing and thus inherits from the capabilities of local message-passing approaches. We show that on both widely used graph benchmarks and synthetic mesh and graph datasets, the proposed framework outperforms state-of-the-art methods by a significant margin.+
Terahertz orbital angular momentum (THz-OAM) technology can greatly improve the communication capacity and is considered to be one of the key technologies for next-generation communication. The rapid and accurate desi...
详细信息
ISBN:
(纸本)9798350389968
Terahertz orbital angular momentum (THz-OAM) technology can greatly improve the communication capacity and is considered to be one of the key technologies for next-generation communication. The rapid and accurate design of THz-OAM Metasurfaces (MS) is a very important technology for the development of communication system. In this paper, we propose a THz-OAM MS based on machinelearning. Taking the element size as the input of the neural network and the phase characteristic as the output, the complex numerical calculation process is omitted. The phase of the required element can be quickly obtained through training, and the THz-OAM MS that can generate vortex waves can be designed according to the orbital angular momentum theory. The simulation results are in good agreement with the design, which verifies the performance of machinelearning, and the method has great potential for the development of future communication system.
breast cancer ranks first in female malignant tumors. Early detection and diagnosis is the key to treatment. This paper uses the open-source load_break_cancer breast cancer data set, mainly uses random forest, support...
详细信息
ISBN:
(纸本)9798400709760
breast cancer ranks first in female malignant tumors. Early detection and diagnosis is the key to treatment. This paper uses the open-source load_break_cancer breast cancer data set, mainly uses random forest, support vector machine, logical regression, Gauss naive Bayesian algorithm, BP neural network algorithm, k-neighborhood algorithm and XGBoost algorithm to classify and predict the breast cancer data set, conducts a lot of training and testing on the data set under a variety of machinelearning algorithms, analyzes the learning curve in the training process, analyzes the training and testing results, and analyzes the performance of the algorithm processing data, which is of great significance for breast cancer diagnosis and treatment.
machine Translation (MT) has made significant advancements due to the rapid expansion of natural language processing. Even though several machine translation methods have been released in recent years, there hasn'...
详细信息
machine Translation (MT) has made significant advancements due to the rapid expansion of natural language processing. Even though several machine translation methods have been released in recent years, there hasn't been enough focus on automated and intelligent quality detection for translation outcomes. Neural machine Translation (NMT), a machine translation method that is datadriven, is more effective in large corpora but with restricted corpus resources, there is still a sizable range of opportunity for advancement. To overcome these issues, design a system module of sentence translator using Neural machine Translation is presented. This analysis will construct a translation system depending on the GRNN (Gated Recurrent Neural network) deep learning algorithm and comprises the creation of the attention, pre-processing, coding, and decoding modules. This system does the Language translation with improved Accuracy and performance.
暂无评论