检索结果-内蒙古大学图书馆

International Conference on Machine Learning and Applications (ICMLA)

作者： Biao Yin Nicholas Josselyn Thomas Considine John Kelley Berend Rinderspacher Robert Jensen James Snyder Ziming Zhang Elke Rundensteiner Department of Data Science Worcester Polytechnic Institute Worcester MA Department of Computer Science Worcester Polytechnic Institute Worcester MA Department of Electrical and Computer Engineering Worcester Polytechnic Institute Worcester MA

Corrosion is a prevalent issue in numerous industrial fields, causing expenses nearing $3 trillion or 4% of the GDP annually with safety threats and environmental pollution. To timely qualify and validate new corrosion-inhibiting materials on a large scale, accurate and efficient corrosion assessment is crucial. Yet it is hindered by a lack of automatic tools for expert-level corrosion segmentation of material science experimental images. Developing such tools is challenging due to limited domain-valid data, image artifacts visually similar to corrosion, various corrosion morphology, strong class imbalance, and millimeter-precision corrosion boundaries. To help the community address these challenges, we curate the first expert-level segmentation annotations for a real-world image dataset [1] for scientific corrosion segmentation. In addition, we design a deep learning based model, called DeepSC-Edge that achieves guidance of ground-truth edge learning by adopting a novel loss that avoids over-fitting to edges. It also is enriched by integrating a class-balanced loss that improves segmentation with small area but crucial edges of interest for scientific corrosion assessment. Our dataset and methods pave the way to advanced deep-learning models for corrosion assessment and generation – promoting new research to connect computer vision and material science discovery. Once the appropriate approvals have been cleared, we expect to release the code and data at: https://***/

关键词：

来源：评论

学校读者我要写书评

暂无评论

Exploiting Interdisciplinary Research Design for Temporally Complex Big data: Discussion of a Case-Study Using on Heterogenous Bibliographic Big data

引用

Proceedings of the Association for Information science and Technology 2022年第1期59卷 579-582页

作者： Mabry Patricia, L. Brian, Martinson Thomas, Valente Liu, Xiaozhong Research Dept. HealthPartners Institute United States Dept of Medicine University of Minnesota United States University of Southern California Irvine United States Data Science & Computer Science Worcester Polytechnic Institute United States

Artificial Intelligence (AI) methods are valued for their ability to predict outcomes from dynamically complex data. Despite this virtue, AI is widely criticized as a "black box" i.e., lacking mechanistic explanations to accompany predictions. We introduce a novel interdisciplinary approach that balances the predictive power of data-driven methods with theory-driven explanatory power by presenting a shared use case from four disciplinary perspectives. The use case examines scientific career trajectories through temporally complex, heterogeneous bibliographic big data. Topics addressed include: data representation in complex problems, trade-offs between theoretical, hypothesis-driven, and data-driven approaches, AI trustworthiness, model fairness, algorithm explainability and AI adoption/usability. Panelists and audience members will be prompted to discuss the value of approach presented versus other ways to address the challenges raised by the panel, and to consider their limitations and remaining challenges. 85 Annual Meeting of the Association for Information science & Technology | Oct. 29 – Nov. 1, 2022 | Pittsburgh, PA. Author(s) retain copyright, but ASIS&T receives an exclusive publication license.

关键词： Big data

来源：评论

学校读者我要写书评

暂无评论

Hybrid Shunted Transformer embedding UNet for remote sensing image semantic segmentation

引用

Neural Computing and Applications 2024年第25期36卷 15705-15720页

作者： Zhou, Huacong Xiao, Xiangling Li, Huihui Liu, Xiaoyong Liang, Peng School of Computer Science Guangdong Polytechnic Normal University Guangdong Guangzhou510630 China Guangdong Provincial Key Laboratory of Intellectual Property and Big Data Guangdong Polytechnic Normal University Guangdong Guangzhou510665 China School of Data Science and Engineering Guangdong Polytechnic Normal University Guangdong Guangzhou510630 China Academy of Heyuan Guangdong Polytechnic Normal University Guangdong Heyuan517099 China The Cyberspace Institute of Advanced Technology Guangzhou University Guangdong Guangzhou510006 China

With the development of deep learning, Remote Sensing Image (RSI) semantic segmentation has produced significant advances. However, due to the sparse distribution of the objects and the high similarity between classes, the task of semantic segmentation in RSI is still extremely challenging. In this paper, we propose a novel semantic segmentation framework for RSI called HST-UNet that can overcome the shortcomings of the existing models and extract and recover the global and local features of RSI, which is a hybrid semantic segmentation model with Shunted Transformer as encoder and Multi-Scale Convolutional Attention Network (MSCAN) as decoder. Then, to better fuse the information from the Encoder and the Decoder and alleviate the ambiguity, we design a Learnable Weighted Fusion (LWF) module to effectively connect to the decoder features. Extensive experiments demonstrate that the proposed HST-UNet outperforms the state-of-the-art methods, achieving F1 score/MIoU accuracy of 71.44%/83.00% on the ISPRS Vaihingen dataset and 77.36%/87.09% on ISPRS Potsdam dataset. The code will be available at https://***/HC-Zhou/HST-UNet. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Detecting the Rumor Patterns Integrating Features of User, Content, and the Spreading Structure 19th

Detecting the Rumor Patterns Integrating Features of User,...

引用

19th International Conference on Wisdom, Well-Being, Win-Win, iConference 2024

作者： Yan, Pengwei Yu, Guo Jiang, Zhuoren Lin, Tianqianjin Yuan, Weikang Liu, Xiaozhong Department of Information Resources Management Zhejiang University Hangzhou310058 China Department of Big Data Security and Secrecy Wuhan University Wuhan430072 China Computer Science Department Worcester Polytechnic Institute WorcesterMA01609-2280 United States

ISBN: (纸本)9783031578595

The openness characteristic of social networks facilitates the rapid spread of rumors, necessitating effective methods for detecting and managing the abundance of rumors on social media. Existing studies have primarily focused on improving the accuracy of rumor detection, but often overlook the vital aspects of interpretability and explanation of rumor patterns, limiting their credibility and real-world usability. Additionally, previous works have typically examined only a subset of user features, content, and spreading structure, neglecting the analysis of compound rules. To address these limitations, we propose a novel framework for detecting rumor patterns that emphasize comprehensive feature construction and the explanation of compound rules. Our framework incorporates multi-dimensional features, including user characteristics, post content, and the structure of information propagation. Advanced techniques, including large language models (such as ChatGPT) and graph motif discovery algorithms, are employed for feature construction. By leveraging diverse features, crucial integrated rules identified by Rulefit can investigate the contextually dependent associations among various interrelated rumor factors. We consolidate and analyze seven distinct rumor patterns based on the Credible Early Detection dataset, deriving valuable insights into the inherent characteristics of rumors. The recognition of rumor patterns empowers social media platforms and fact-checking organizations to develop targeted and explainable interventions that effectively mitigate the spread of rumors and safeguard the integrity of information. These interventions greatly enhance the transparency and trustworthiness of rumor management, fostering a more reliable information ecosystem. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

关键词： Information dissemination

来源：评论

学校读者我要写书评

暂无评论

Feasibility of Spectra-Based Quantification Algorithm for Non-Linear Photoacoustic Contrast Agents

Feasibility of Spectra-Based Quantification Algorithm for No...

引用

Photons Plus Ultrasound: Imaging and Sensing 2023

作者： Serebrennikova, Liudmila Gao, Shang Murakami, Ryo Boinapally, Srikanth Ray, Sangeeta Pomper, Martin G. Zhang, Haichong K. Biomedical Engineering Department Worcester Polytechnic Institute MA United States Robotics Engineering Department Worcester Polytechnic Institute MA United States Computer Science Department Worcester Polytechnic Institute MA United States Radiology and Radiological Science Johns Hopkins University School of Medicine ME United States

ISBN: (纸本)9781510658639

Photoacoustic (PA) imaging is a technique that visualizes the optical absorption characteristics with ultrasound-like spatial resolution. It has been demonstrated that molecular targeted contrast agents (CAs) have great potential in applying to the early detection of tumor. These CAs attach to the tissues of the subject of interest, i.e., the cancerous cells, and emit PA signals based on the absorbed light. Precise quantification of the CA expression is desired to identify the degree of the malignancies. The current quantification of the CA expression is based on the strength of the PA intensity with the assumption of its linearity between them. However, the estimation accuracy of such a method is limited because the PA intensity is affected by many factors including the light, sound, and tissue interactions and the use of CAs that do not present linearity. Here, we investigate a robust quantification method by using the spectroscopic relationship between the dye concentration and its corresponding PA variation. We introduce a spectroscopic decomposition algorithm considering multiple reference spectra to accommodate the non-linear behavior of sample concentrations. The in vitro validation of the concept was performed using the synthesized PA CA possessing the non-linear property. The concentration of the samples was successfully estimated with the algorithm. The introduced method improved the quantification accuracy by reducing the averaging estimation error from 15.89 μM to 1.80 μM, compared with the conventional estimation. © 2023 SPIE.

关键词： Tissue

来源：评论

学校读者我要写书评

暂无评论

Screening for suicidal ideation with text messages

Screening for suicidal ideation with text messages

引用

2021 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2021

作者： Tlachac, M.L. Dixon-Gordon, Katherine Rundensteiner, Elke Data Science Worcester Polytechnic Institute Psychological and Brain Sciences University of Massachusetts Amherst Data Science and Computer Science Worcester Polytechnic Institute

ISBN: (纸本)9781665403580

Suicide is a leading cause of death in the US, with suicide rates increasing annually. Passive screening of suicidal ideation is vital to provide referrals to at-risk individuals. We study to what degree smartphone-based communication, in particular, text messages, could be leveraged for passively screening for suicidal ideation. We analyze the screening ability of texts sent in different time periods prior to reported ideation, namely, texts from specific weeks only versus accumulative over several weeks. Our approach involves performing comprehensive feature engineering and identifying influential features to train machine learning models. With just the prior week of texts, we were able to predict the existence of suicidal ideation with AUC = 0.88, F1 = 0.84, accuracy = 0.81, sensitivity = 0.94, and specificity = 0.68. The most influential features include word frequencies of words in the car, clothing, affection, confusion, driving, real estate, and journalism categories. This research, demonstrating the potential of text messages to screen for suicidal ideation, will guide the development of screening technologies. © 2021 IEEE

关键词： mHealth

来源：评论

学校读者我要写书评

暂无评论

Improved IChOA-Based Reinforcement Learning for Secrecy Rate Optimization in Smart Grid Communications

引用

computers, Materials & Continua 2024年第11期81卷 2819-2843页

作者： Mehrdad Shoeibi Mohammad Mehdi Sharifi Nevisi Sarvenaz Sadat Khatami Diego Martín Sepehr Soltani Sina Aghakhani The WPI Business School Worcester Polytechnic InstituteWorcesterMA 01609-2280USA Department of Computer Science Escuela de Ingeniería Informática de SegoviaUniversidad de ValladolidSegovia40005Spain Department of Data Science Engineering University of HoustonHoustonTX 77204USA Department of Industrial Engineering College of EngineeringUniversity of HoustonHoustonTX 77204USA Department of Industrial and Manufacturing Systems Engineering Iowa State UniversityAmesIA 50011USA

In the evolving landscape of the smart grid(SG),the integration of non-organic multiple access(NOMA)technology has emerged as a pivotal strategy for enhancing spectral efficiency and energy ***,the open nature of wireless channels in SG raises significant concerns regarding the confidentiality of critical control messages,especially when broadcasted from a neighborhood gateway(NG)to smart meters(SMs).This paper introduces a novel approach based on reinforcement learning(RL)to fortify the performance of *** by the need for efficient and effective training of the fully connected layers in the RL network,we employ an improved chimp optimization algorithm(IChOA)to update the parameters of the *** integrating the IChOA into the training process,the RL agent is expected to learn more robust policies faster and with better convergence properties compared to standard optimization *** can lead to improved performance in complex SG environments,where the agent must make decisions that enhance the security and efficiency of the *** compared the performance of our proposed method(IChOA-RL)with several state-of-the-art machine learning(ML)algorithms,including recurrent neural network(RNN),long short-term memory(LSTM),K-nearest neighbors(KNN),support vector machine(SVM),improved crow search algorithm(I-CSA),and grey wolf optimizer(GWO).Extensive simulations demonstrate the efficacy of our approach compared to the related works,showcasing significant improvements in secrecy capacity rates under various network *** proposed IChOA-RL exhibits superior performance compared to other algorithms in various aspects,including the scalability of the NOMA communication system,accuracy,coefficient of determination(R2),root mean square error(RMSE),and convergence *** our dataset,the IChOA-RL architecture achieved coefficient of determination of 95.77%and accuracy of 97.41%in validation *** was accompanied by the lowes

关键词： Smart grid communication secrecy rate optimization reinforcement learning improved chimp optimization algorithm

来源：评论

学校读者我要写书评

暂无评论

Semantically Encoding Activity Labels for Context-Aware Human Activity Recognition

Semantically Encoding Activity Labels for Context-Aware Huma...

引用

IEEE International Conference on Pervasive Computing and Communications (PerCom)

作者： Wen Ge Guanyi Mou Emmanuel O. Agu Kyumin Lee Computer Science Department Worcester Polytechnic Institute Worcester MA USA

ISBN: (数字)9798331535513

ISBN: (纸本)9798331535520

Prior work has primarily formulated Context-Aware Human Activity Recognition (CA-HAR) as a multi-label classification problem, where model inputs are time-series sensor data and target labels are binary encodings representing whether a given activity or context occurs. These CA-HAR methods either predicted each label independently or manually imposed relationships using graphs. However, both strategies often neglect an essential aspect: activity labels have rich semantic relationships. For instance, walking, jogging, and running activities share similar movement patterns but differ in pace and intensity, indicating that they are semantically related. Consequently, prior CA-HAR methods often struggled to accurately capture these inherent and nuanced relationships, particularly on datasets with noisy labels typically used for CA-HAR or situations where the ideal sensor type is unavailable (e.g., recognizing speech without audio sensors). To address this limitation, we propose Semantically Encoding Activity Labels (SEAL), which leverage Language Models (LMs) to encode CA-HAR activity labels to capture semantic relationships. LMs generate vector embeddings that preserve rich semantic information from natural language. Our SEAL approach encodes input-time series sensor data from smart devices and their associated activity and context labels (text) as vector embeddings. During training, SEAL aligns the sensor data representations with their corresponding activity/context label embeddings in a shared embedding space. At inference time, SEAL performs a similarity search, returning the CA-HAR label with the embedding representation closest to the input data. Although LMs have been widely explored in other domains, surprisingly, their potential in CA-HAR has been underexplored, making our approach a novel contribution to the field. Our SEAL approach has been rigorously evaluated on three real-world datasets, demonstrating its superior performance. It consistently outperforms

关键词： Training Semantics Seals Speech recognition Encoding data models Vectors Human activity recognition Smart devices Context modeling

来源：评论

学校读者我要写书评

暂无评论

Soil Subsurface Channel Statistical Characterization for Drone-Borne Intelligent GPR Advancement

Soil Subsurface Channel Statistical Characterization for Dro...

引用

IEEE International Conference on Wireless for Space and Extreme Environments (WiSEE)

作者： Noushin Khosravi Largani Seyed Reza Zekavat Vincent Filardi Data Science Department Worcester Polytechnic Institute (WPI) Worcester MA USA

ISBN: (数字)9798350351118

ISBN: (纸本)9798350351125

The advancement of drone-borne intelligent Ground Penetration Radar (GPR) is hinged upon accurate received signal feature extraction, which relies on high-performance soil subsurface channel estimation. This requires the optimal design of the transmitted waveform, which needs soil subsurface channel parameter statistics. In this paper, we investigate the statistics of soil subsurface channel parameters for different soil textural classes (defined by sand and clay percentage, and adopted based on the USDA textural triangle). Channel parameters statistics of interest are the gain correlation matrices, the statistical distributions of gain and Time-of-Arrival (ToA), and the mean Power-Delay Profile (PDP). Channel gain correlation matrix and its distribution are critical to waveform design. Mean PDP offers insight into the number of soil layers. The results of this paper contribute to the intelligent optimal waveform design for GPR, the implementation of feature extraction in ML applications, and accurate estimation of soil layers, moisture, and permittivity.

关键词： Wireless communication Maximum likelihood estimation Correlation Accuracy Moisture Channel estimation Soil Feature extraction Soil texture Permittivity

来源：评论

学校读者我要写书评

暂无评论

Microservice Architecture of Modern eLearning Application 5

Microservice Architecture of Modern eLearning Application

引用

5th IEEE International Conference on Communications, Information, Electronic and Energy Systems, CIEES 2024

作者： Schukin, Alexander Scerbakov, Nikolai Rezedinova, Eugenia Peter the Great St. Petersburg Polytechnic University Institute of Computer Science and Cybersecurity St. Petersburg Russia Graz University of Technology Institute of Interactive Systems and Data Science Graz Austria

ISBN: (纸本)9798350352863

Learning Management Systems (LMS) have been integral to educational institutions for over 20 years. Technically, an LMS consists of functional components designed to deliver, discuss, explain educational materials, and assess user knowledge through various paradigms such as uploading assignments and online questionnaires. The prolonged and intensive use of such systems emphasizes the importance of their maintainability, allowing the accommodation of new components and the replacement of outdated ones. Over the years, the technological foundation of LMS components has transitioned from early versions of PHP to advanced Java, Python, and JavaScript packages. data repositories have evolved from local relational databases to shared NoSQL databases. Communication technologies have shifted from NNTP news servers to multiple social applications like WhatsApp and *** this paper, we present architectural solutions implemented in a modern LMS called WBT-Master. This architecture specifically addresses the maintainability of modern LMS, ensuring they evolve alongside emerging Internet technologies. The practical significance of our work lies in its ability to support very heterogeneous components within a logically unified LMS framework. By discussing the architecture of WBT-Master, we offer a method for incorporating both old and new technological platforms into a single system. Additionally, we use this architecture to interface with cloud services, foster social interactions, and optimize user experience. This contributes to the evolution of E-Learning platforms, offering enhanced and more interactive educational experiences for both learners and educators. © 2024 IEEE.

关键词： Cloud computing architecture

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：