this paper proposes a novel optimization method designed to address the challenges of training deep neural networks on imbalanced datasets. the main contributions include the proposal of a novel trajectory-unified met...
详细信息
ISBN:
(数字)9798331542856
ISBN:
(纸本)9798331542863
this paper proposes a novel optimization method designed to address the challenges of training deep neural networks on imbalanced datasets. the main contributions include the proposal of a novel trajectory-unified method based on the influence balance factor, which adaptively reweights samples to alleviate the overfitting of decision boundaries to the majority class. In this paper a rigorous theoretical framework is presented that validates the stability and convergence properties of the proposed method. Extensive experiments demonstrate that this method outperforms cutting-edge techniques in handling class imbalance issues. In summary, this approach effectively tackles the challenges posed by imbalanced data distribution, providing a systematically enhanced solution to the class imbalance problem in advanced deep learning models.
Recent object detection developments have been largely influenced by advances in the field of deep learning. In this context, we have developed a pallet detection model integrated into a forklift robot in order to opt...
详细信息
ISBN:
(数字)9798350353839
ISBN:
(纸本)9798350353846
Recent object detection developments have been largely influenced by advances in the field of deep learning. In this context, we have developed a pallet detection model integrated into a forklift robot in order to optimize the picking process in storage and logistics environments in industrial settings. the project based on the architecture of the YOLO (You Only Look Once) model to improve pallet recognition and estimation of position and distance allowing robots to autonomously locate and pick pallets. this model stands out for its ability to provide fast and accurate object detection in high-resolution images, making them particularly suitable for real-time applications such as robotics and logistics.
PV power prediction based on historical data and deep learning has been widely used in power systems to improve prediction accuracy. However, it is difficult to establish an accurate PV power prediction model for newl...
PV power prediction based on historical data and deep learning has been widely used in power systems to improve prediction accuracy. However, it is difficult to establish an accurate PV power prediction model for newly constructed PV power plants due to the lack of historical data. therefore, based on the migration learning framework, this paper proposes a long and short-term memory network (CNN-LSTM) withthe fusion of VMD-CNN-LSTM based on the variable mode decomposition (VMD). therefore, based on the transfer learning framework, this paper proposes a VMD-CNN-LSTM method for PV power prediction based on the convergence of variational mode decomposition (VMD) and convolutional neural network short term memory (CNN-LSTM). Firstly, VMD is used to decompose the preprocessed historical PV power and meteorological data to reduce its complexity and random volatility; secondly, the decomposed sequences are predicted using a combined CNN-LSTM model, and the prediction results of each component are integrated; finally, Transfer learning (TL) is introduced to fine-tune the constructed CNN-LSTM model. through the way of model parameter sharing, the pre-trained model is extrapolated to the data-scarce target PV power plant to realize the migration learning of data features under the condition of few samples. the method is used to conduct experiments on PV power plants in data-poor areas, and the results show that the algorithm proposed in this paper can achieve higher accuracy compared with other prediction models.
Cybersecurity is a significant concern for organisations in the modern, globally interconnected world, but it's also a dynamic field. Conventional security measures are often insufficient due to the growing comple...
Cybersecurity is a significant concern for organisations in the modern, globally interconnected world, but it's also a dynamic field. Conventional security measures are often insufficient due to the growing complexity of attackers. this research investigates the potential for enhancing cybersecurity protocols in business operations via the integration of machine learning methods. In order to emphasise the urgent need for creative solutions, we first provide a high-level summary of the present state of cyber threats. In the first segment, machine learning is introduced along with its application to cybersecurity, with a focus on the need for adaptable models and data-driven insights. the various machine learning methods that may be used to detect and prevent cyberattacks are then covered. these algorithms include those for identifying patterns in data, evaluating behaviour, and deciphering natural language. the study then moves on to real-world commercial applications, such as threat intelligence, network security, and user behaviour analysis. We demonstrate how machine learning may enable businesses to proactively detect and address security breaches, resulting in cybersecurity plans that are more adaptable and durable. Finally, we discuss the difficulties and constraints associated with using machine learning to cybersecurity, along withthe moral issues pertaining to data protection and openness.
Satellite imagery has become widely available for detecting patterns by applying different deep learning and machine learning techniques to it. this paper will cover water monitoring, land cover and tree cover visuali...
详细信息
the use and significance of Artificial Intelligence (AI) are widely discussed across various fields, including business and science. AI, a key branch of computer science, enables algorithms to perform tasks that tradi...
详细信息
the proceedings contain 15 papers presendted at a virtual meeting. the special focus in this conference is on Software Architecture Erosion and Architectural Consistency. the topics include: Interactive Elicitation of...
ISBN:
(纸本)9783031151156
the proceedings contain 15 papers presendted at a virtual meeting. the special focus in this conference is on Software Architecture Erosion and Architectural Consistency. the topics include: Interactive Elicitation of Resilience Scenarios Based on Hazard Analysis Techniques;Towards an Extensible Approach for Generative Microservice Development and Deployment Using LEMMA;applying Knowledge-Driven Architecture Composition with Gabble;architectural optimization for Confidentiality Under Structural Uncertainty;foundations and Research Agenda for Simulation of Smart Ecosystems Architectures;Blended Graphical and Textual Modelling of UML-RT State-Machines: An Industrial Experience;toward Awareness Creation of Common Challenges Women are Facing in Academia: A Study from a German Perspective;mapping Source Code to Modular Architectures Using Keywords;hierarchical Code-to-Architecture Mapping;Building the MSR Tool Kaiaulu: Design Principles and Experiences;self-adaptive Machine learning Systems: Research Challenges and Opportunities;behavioral Maps: Identifying Architectural Smells in Self-adaptive Systems at Runtime;an Architectural Approach for Enabling and Developing Cooperative Behaviour in Diverse Autonomous Robots.
there are two ways to optimize the performance of machine learning models, one is model centric, and another is data centric. Model centric approaches sometimes cannot yield expected results, so researchers and develo...
详细信息
the widespread popularity of Federated learning (FL) has led researchers to delve into its various facets, primarily focusing on personalization, fair resource allocation, privacy, and global optimization, with less a...
详细信息
ISBN:
(数字)9798350376968
ISBN:
(纸本)9798350376975
the widespread popularity of Federated learning (FL) has led researchers to delve into its various facets, primarily focusing on personalization, fair resource allocation, privacy, and global optimization, with less attention puts towards the crucial aspect of ensuring efficient and cost-optimized communication between the FL server and its agents. A major challenge in achieving successful model training and inference on distributed edge devices lies in optimizing communication costs amid resource constraints, such as limited bandwidth, and selecting efficient agents. In resource-limited FL scenarios, where agents often rely on unstable networks, the transmission of large model weights can substantially degrade model accuracy and increase communication latency between the FL server and agents. Addressing this challenge, we propose a novel strategy that integrates a knowledge distillation technique with a Particle Swarm optimization (PSO)-based FL method. this approach focuses on transmitting model scores instead of weights, significantly reducing communication overhead and enhancing model accuracy in unstable environments. Our method, with potential applications in smart city services and industrial IoT, marks a significant step forward in reducing network communication costs and mitigating accuracy loss, thereby optimizing the communication efficiency between the FL server and its agents.
this study addresses the issues of low energy efficiency and high energy consumption in traditional gravure printing machine drying systems by proposing an optimization approach using transcritical CO 2 heat pump tech...
详细信息
ISBN:
(数字)9798331542917
ISBN:
(纸本)9798331542924
this study addresses the issues of low energy efficiency and high energy consumption in traditional gravure printing machine drying systems by proposing an optimization approach using transcritical CO 2 heat pump technology. through the establishment of a mathematical model for the CO 2 transcritical heat pump system, the influence of key parameters such as inlet air temperature and airflow rate in evaporators and gas coolers on the system's Coefficient of Performance (COP) and outlet air temperature was systematically analyzed. the results demonstrate that when the evaporator inlet air temperature increases from 20°C to 40°C, the system COP improves from 3.489 to 4.096. Similarly, increasing the evaporator airflow rate from 2000 mvh to 4000 m 3 /h elevates the COP to 4.096. Conversely, raising the gas cooler inlet air temperature from 31°C to 50°C causes a significant COP decline from 4.096 to 1.808. Furthermore, enhancing the gas cooler airflow rate effectively improves COP. these findings indicate that optimizing inlet air parameters can substantially enhance system energy efficiency. the research provides theoretical foundations for energy-efficient design of gravure printing machine drying systems and promotes industrial applications of transcritical CO 2 heat pump technology.
暂无评论