The proceedings contain 79 papers. The topics discussed include: research on enterprise data assets based on fusion neural network algorithm;analysis of non-heritage film and television viewing based on ADABOOST-BP al...
ISBN:
(纸本)9798350352894
The proceedings contain 79 papers. The topics discussed include: research on enterprise data assets based on fusion neural network algorithm;analysis of non-heritage film and television viewing based on ADABOOST-BP algorithm;intelligent speech enhancement algorithm model based on recorded classroom audio;load forecasting for electric propulsion vessels based on RBF neural network;research on power load forecasting based on a time convolutional network-autoencoder stacked extreme learning machine model;research on sentiment analysis method of fan culture based on CNN-LSTM;ship target recognition based on improved residual network;and research on motion posture based on convolutional neural network algorithm.
A prediction method is proposed based on RBF neural network through in-depth study of power load of ships. Before predicting the ship's power load, it is necessary to pre-process various data of the ship, screen a...
详细信息
ISBN:
(纸本)9798350352900;9798350352894
A prediction method is proposed based on RBF neural network through in-depth study of power load of ships. Before predicting the ship's power load, it is necessary to pre-process various data of the ship, screen and clean the abnormal data, and then normalise the normal samples, and finally construct a model to process the data and optimise the model according to the results. Various data of an electric propulsion ship are selected as the input reference of the model, and a prediction model is established using Matlab. The design method is shown to be one of the methods with high prediction accuracy according to the experimental results, and shows high model credibility in prediction.
With the rapid development of the Marine economy in China, the intelligent identification of ships is significant to Marine traffic safety. Conventional machine learning techniques exhibit low accuracy and slow perfor...
详细信息
ISBN:
(纸本)9798350352900;9798350352894
With the rapid development of the Marine economy in China, the intelligent identification of ships is significant to Marine traffic safety. Conventional machine learning techniques exhibit low accuracy and slow performance in image object recognition tasks. In response to this issue, this paper proposes a ship target recognition method based on improved ResNet18. Firstly, a Convolutional Block Attention Module (CBAM) is added after the last convolutional layer in the network to enhance the feature extraction capability for ship targets. Additionally, GELU activation function and dropout are introduced to address overfitting caused by scarce dataset, utilizing data augmentation techniques. Lastly, a comparison of the improved residual model with those of other neural network models is made. The validation set precision of the improved ResNet18 model is 96.39%, which is 6.38%, 5.05%, 4.61% and 3.13% higher than that of AlexNet, VGGNet, GoogLeNet and ResNet18, respectively, according to the experimental results.
In modern aerial warfare, situational awareness plays a critical role in enhancing combat effectiveness, and intent recognition in air combat is the core element for achieving this *** the limitations of traditional a...
详细信息
ISBN:
(纸本)9798350352900;9798350352894
In modern aerial warfare, situational awareness plays a critical role in enhancing combat effectiveness, and intent recognition in air combat is the core element for achieving this *** the limitations of traditional air combat intent recognition methods in complex operational environments, this paper proposes a novel multitask intent recognition method based on feature *** method first enhances the correlation between feature space and sample space through contrastive learning, effectively addressing the challenge of imbalanced aerial target datasets. Subsequently, a hierarchical variable-length Long Short-Term Memory (LSTM) model is employed to extract and process information across different time scales, leading to a more comprehensive and accurate understanding of target *** results indicate that our method demonstrates significantly superior performance in complex air combat environments. It notably enhances recognition stability, accuracy, and speed, showcasing strong potential for practical applications.
This paper presents a comprehensive approach for enterprise data asset processing, based on the Random Forest (RF) algorithm integrated with Back Propagation (BP) neural networks, to address the challenge of large-sca...
详细信息
ISBN:
(纸本)9798350352900;9798350352894
This paper presents a comprehensive approach for enterprise data asset processing, based on the Random Forest (RF) algorithm integrated with Back Propagation (BP) neural networks, to address the challenge of large-scale, high reliability, and real-time requirements for power grid data asset information management that traditional methods are unable to meet. This method not only ensures the reliability and timeliness of data processing, but also enhances the efficiency of data processing, playing a significant role in the scientific and rational exploration of the value of big data for enterprises. By utilizing the Random Forest algorithm classifier to categorize data types of the power grid and inputting them into the BP neural network for analysis and learning, the proposed algorithm, when integrated into the construction of an enterprise digital asset management system, yields corresponding results for system service requirements. Experimental results demonstrate that the processing time of cloud-edge collaborative data is significantly shorter than that of cloud computing, and the classification accuracy of the proposed technology exceeds 92%. Moreover, the load forecasting results exhibit minimal errors, thereby substantiating the effectiveness and reliability of this technology.
The aging population is accompanied by an increase in the incidence of mild cognitive impairment (MCI). MCI is a pre-dementia stage that can be prevented by early intervention. Traditional MCI interventions are typica...
详细信息
ISBN:
(纸本)9798350370027;9798350370034
The aging population is accompanied by an increase in the incidence of mild cognitive impairment (MCI). MCI is a pre-dementia stage that can be prevented by early intervention. Traditional MCI interventions are typically delivered in person at community centers or clinics, which can be difficult to access and travel to. In recent years, there has been growing interest in the use of extended reality (XR) for MCI intervention. XR offers advantages such as remote delivery, immersion, and personalization. This study developed and evaluated an XR-based cognitive and physical training system for older adults and MCI patients. By integrating multimodal devices and comprehensive XR assistive technologies, the system aims to enhance cognitive and physical abilities through personalized, engaging training experiences. The evaluation results showed that the system had high user engagement and usability. Additionally, the system was shown to promote cognitive and physical function improvement. The study findings suggest that XR-based cognitive and physical training systems could be a promising tool for MCI intervention.
Modern engineering practice includes complex decision-making processes that require analyzing huge data and taking into account many factors. cognitivecomputing is a way of analyzing information that can enhance the ...
详细信息
YOLOv5 is a widely used object detection method for its high-performance of balancing speed and accuracy. data transfer for YOLOv5 is a challenging task, as it could involve domain shift between different datasets. We...
详细信息
ISBN:
(纸本)9798350352900;9798350352894
YOLOv5 is a widely used object detection method for its high-performance of balancing speed and accuracy. data transfer for YOLOv5 is a challenging task, as it could involve domain shift between different datasets. We propose a data transfer method for YOLOv5 based on a Selector Network and Partial Pseudo-Labeling. The backbone network of YOLOv5 is used for feature extraction, a Selector Network is employed to filter source domain outlier samples to address the domain shift issue, and Partial Pseudo-Labels are leveraged to improve classification performance in the target domain. The results show that our method improves detection accuracy by an average of 2.4% on public datasets (COCO and VOC) and our datasets, with a notable improvement of 3.1% under complex backgrounds and varying lighting conditions. It demonstrates that our method can effectively enhances the accuracy and stability of YOLOv5.
data activities and technologies in modern organisations are becoming more diverse and complex, and continuously operate, change, and evolve as an enterprise data ecosystem (EDE). This fast-changing environment presen...
详细信息
ISBN:
(纸本)9798350368567;9798350368550
data activities and technologies in modern organisations are becoming more diverse and complex, and continuously operate, change, and evolve as an enterprise data ecosystem (EDE). This fast-changing environment presents both challenges and opportunities for many computing and design concepts and approaches and raises the questions whether and how they can be continuously and effectively applied. This paper analyses operation features and complexity of EDE and delves into design considerations for data products and data-as-a-Service (DaaS) to play their roles as solutions for data practice management improvement, integration, interoperability and design of services in data space.
With the advancement of society and economy, projects are becoming larger and more complex, requiring project managers to have a higher level of expertise. The emergence of big data technology presents both opportunit...
详细信息
ISBN:
(纸本)9798350391961;9798350391954
With the advancement of society and economy, projects are becoming larger and more complex, requiring project managers to have a higher level of expertise. The emergence of big data technology presents both opportunities and challenges for project management. By effectively managing a variety of information, project delivery times can be reduced, risks can be mitigated in a scientific manner, and project schedules can be optimized. This article explores the use of big data in project management, focusing on schedule control and optimization analysis, offering fresh perspectives for innovation in project management. In today's complex project environment, it is essential to fully leverage technology to improve project management skills and overcome the challenges posed by the abundance of data.
暂无评论