The Underwater Internet of Things (IoUT) shows great potential in enabling a smart ocean. Underwater sensor network (UWSN) is the main form of IoUT, but it faces the problem of reliable data transmission. To address t...
The Underwater Internet of Things (IoUT) shows great potential in enabling a smart ocean. Underwater sensor network (UWSN) is the main form of IoUT, but it faces the problem of reliable data transmission. To address these issues, this paper considers the use of autonomous underwater vehicles (AUV) as mobile collectors to build a reliable dynamic data collection system, while using Value of Information (VoI) as a primary metric to measure data quality. This paper first builds a realistic model to characterize the behavior of AUV and sensor nodes and challenging environments. Then a method based on deep reinforcement learning is used to dynamically plan the AUV's navigation route by jointly considering the location of nodes, the value of node data information and the state of AUV, with the goal of maximizing the data collection efficiency of AUV. The simulation results show that the dynamic data collection scheme is superior to the traditional path planning scheme which only considers the node location, and can greatly improve the efficiency of AUV data collection.
The proceedings contain 39 papers. The special focus in this conference is on Artificial Intelligence and Pattern Recognition. The topics include: Echo State Networks for the Prediction of Chaotic Syste...
ISBN:
(纸本)9783031495519
The proceedings contain 39 papers. The special focus in this conference is on Artificial Intelligence and Pattern Recognition. The topics include: Echo State Networks for the Prediction of Chaotic Systems;weighted t-Distributed Stochastic Neighbor Embedding for Projection-Based Clustering;Evaluation of XAI Methods in a FinTech Context;towards Automatic Principles of Persuasion Detection Using Machine Learning Approach;indirect Condition Monitoring of the Transmission Belts in a Desalination Plant by Using Deep Learning;a Novel Method for Filtering a Useful Subset of Composite Linguistic Summaries;harnessing Key Phrases in Constructing a Concept-Based Semantic Representation of Text Using Clustering Techniques;a Comparative Study of Deep Learning Methods for Brain Magnetic Resonance Image Reconstruction;enhancing Spanish Aspect-Based Sentiment Analysis Through Deep Learning Approach;oversampling Method Based Covariance Matrix Estimation in High-Dimensional Imbalanced Classification;a Hybrid Approach for Spanish Emotion Recognition Applying Fuzzy Semantic processing;a Knowledge-Based User Feedback Classification Approach for Software Support;A Metaphorical Text Classifier to Compare the Use of RoBERTa-Large, RoBERTa-Base and BERT-Base Uncased;Improvements to the IntiGIS Model Related to the Clustering of Consumers for Rural Electrification;diagnosis of Alzheimer Disease Progression Stage from Cross Sectional Cognitive data by Deep Neural Network;Robust MCU Oriented KWS Model for Children Robotic Prosthetic Hand Control;detection of Malicious Bots Using a Proactive Supervised Classification Approach;hybrid Selection of Breast Cancer Risk Factors in Cuban Patients;good Negative Sampling for Triple Classification;polarity Prediction in Tourism Cuban Reviews Using Transformer with Estimation of Distribution algorithms;multivariate Cuban Consumer Price Index database, Statistic Analysis and Forecast Baseline Based on Vector Autoregressive;SqueezerFaceNet: Reducing a Small
In this work, aconcentric elliptical array of antennas is optimized with the help of three relatively new and efficient meta-heuristic algorithms; namely, Teaching Learning Based Optimization (TLBO), Multi-Verse Optim...
In this work, aconcentric elliptical array of antennas is optimized with the help of three relatively new and efficient meta-heuristic algorithms; namely, Teaching Learning Based Optimization (TLBO), Multi-Verse Optimization (MVO), and Grey Wolf Optimization(GWO). This work demonstrates a detailed study of the effect of various antenna array contro lvariables for the improvemen to normalized power patterns generated by elliptical array *** objective of this study is to reduce Side Lobe Level (SLL) value to the possible amount and simultaneously reach the desired First Null BeamWidth (FNBW) value. A number of cases are studied and the outcomes are exhibited in a desegregated way with the best optimal values of eccentricities, semi-major axis, and inter-element spacing. An investigation is performed to explore theperformance of the algorithms based on the common antenna parameters such as the number of elements, obtained SLL, and obtained FNBW. Additionally, the statistical specifications of all the algorithms are examined in terms of mean, standard deviation, best objective function, and time complexity. T-test iscarried out at the end of this study, separately on each acquired data set to authenticate the result.
Big data is the present era in which we find ourselves. Using spatial data, including auxiliary geospatial datasets and remotely sensed satellite images, has become prevalent in land cover and land use mapping (LCLU)....
Big data is the present era in which we find ourselves. Using spatial data, including auxiliary geospatial datasets and remotely sensed satellite images, has become prevalent in land cover and land use mapping (LCLU). Moreover, deep learning and machine learning algorithms have recently been innovated, offering new prospects for LCLU mapping. But problems often arise when it comes to using big geospatial data. This article summarizes research advancements in remote sensing (RS), machine learning (ML), deep learning (DL), and geographic information big data for LCLU classification. We analyzed the advantages, disadvantages, and potential future directions of LCLU mapping utilizing big geospatial data. Further research is needed to improve the LCLU process at larger scales.
Increasing temperatures regularly cause dry vegetation in wooded regions to burn due to sun radiation, which in turn sparks forest fires. Predicting and mapping fire susceptibility in fire-prone forests worldwide is a...
Increasing temperatures regularly cause dry vegetation in wooded regions to burn due to sun radiation, which in turn sparks forest fires. Predicting and mapping fire susceptibility in fire-prone forests worldwide is a critical research focus. This paper presents a data mining approach for forest fire prediction. The maps that arise from evaluating the Gini index's performance in identifying significant features serve as a basis for firefighting tactics and regulations that are more successful in promoting sustainable forest resource management. The proposed model utilizes Support Vector Machine for regression and achieves an improved accuracy of approximately 87% compared to previous research. This study emphasizes the importance of enhancing traditional algorithms by prioritizing key features to ensure consistent results and optimize predictive models.
Multi-node intelligent of power material supply chain under green and low-carbon goals with AI simulation is studied. This study combines the supply chain operation performance evaluation with the enterprise process m...
详细信息
Multi-node intelligent of power material supply chain under green and low-carbon goals with AI simulation is studied. This study combines the supply chain operation performance evaluation with the enterprise process monitoring system to fully mobilize the physical support role of the data center is considered and then form a comprehensive operation performance evaluation system. The multi-node intelligent of power material supply chain is implemented with the KNN and the data mining algorithms. The models are designed based on the multi-core cluster systems with different node computing, communication and storage capabilities, the Kubernetes is selected as the model, and VMI is selected as the tool to finish the work of bad data identification. Through the verification on the AI simulation, the performance is validated. It can be seen from the experiment that the model error is between [0], [6], which belongs to the low or medium levels which are acceptable.
As new ways to generate renewable energy cost effectively are being developed, high load devices such as electric vehicles (EVs) are becoming common. EVs require high current and voltage for charging, due to the large...
详细信息
We introduce a method for automatic corrosion detection based on the application of machine learning techniques to 3D point cloud data generated by a LIDAR sensor. In our approach a point is assigned one of the consid...
详细信息
Brain tumors are irregular growths of cells within the brain, and some of them can be cancerous. MRI is the typical diagnostic tool for identifying brain tumors. MRI images can reveal any anomalous tissue regeneration...
Brain tumors are irregular growths of cells within the brain, and some of them can be cancerous. MRI is the typical diagnostic tool for identifying brain tumors. MRI images can reveal any anomalous tissue regeneration in the brain. Applying novel algorithms to MRI images, predicting a tumor becomes much faster and more accurate to help deliver treatments to patients. These predictions also help radiologists make quick decisions. In the proposed work to enhance diversity in the dataset, data augmentation techniques are utilized. The presence of brain tumors is detected using VGG19 and InceptionV3 neural networks, with the performance visualized through Grad CAM.
Aiming at the problem of inconsistent charging standards and restricted import and export of electric vehicles in various countries today, this essay propose a Chinese-European Standard Protocol Converter for Charging...
详细信息
Aiming at the problem of inconsistent charging standards and restricted import and export of electric vehicles in various countries today, this essay propose a Chinese-European Standard Protocol Converter for Charging Piles, which enables two-way communication between GB/T 27930 digital communication protocol for Chinese standard piles and ISO 15118 digital communication protocol for European standard vehicles. The converter is based on commercially available charging conversion modules, with the STM32F413 chip and the new highly integrated MSE102X carrier chip, to build a series of sub-functional module circuits, providing new ideas for future updates to the charging converter. This device offers great dataprocessing capability, low power consumption, and high charging efficiency.
暂无评论