检索结果-内蒙古大学图书馆

IEEE Transactions on Artificial Intelligence 2023年第1期4卷 81-91页

作者： Mishra, Aakansha Anand, Ashish Guha, Prithwijit Department of Computer Science & Engineering Guwahati781039 India Department of Electronics and Electrical Engineering Guwahati781039 India Mehta Family School of Data Science and Artificial Intelligence Indian Institute of Technology Guwahati Guwahati781039 India

Visual question answering (VQA) aims at predicting an answer to a natural language question associated with an image. This work focuses on two important issues pertaining to VQA, which is a complex multimodal AI task: First, the task of answer prediction in a large output answer space, and second, to obtain enriched representation through cross-modality interactions. This work aims to address these two issues by proposing a dual attention (DA) and question categorization (QC)-based visual question answering model (DAQC-VQA). DAQC-VQA has three main network modules: First, a novel dual attention mechanism that helps toward the objective of obtaining an enriched cross-domain representation of the two modalities;second, a question classifier subsystem for identifying input (natural language) question category. The second module of question categorizer helps in reducing the answer search space;and third, a subsystem for predicting answer depending on the question category. All component networks of DAQC-VQA are trained in an end-to-end manner with a joint loss function. The performance of DAQC-VQA is evaluated on two widely used VQA datasets, viz., TDIUC and VQA2.0. Experimental results demonstrate competitive performance of DAQC-VQA against the recent state-of-the-art VQA models. An ablation analysis indicates that the enriched representation obtained using the proposed dual-attention mechanism helps improve performance. © 2022 IEEE.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

A Secure Approach to Avoid data Repetition in Cloud Storage Systems 2

A Secure Approach to Avoid Data Repetition in Cloud Storage ...

引用

2nd International Conference on Artificial Intelligence and Machine Learning Applications, AIMLA 2024

作者： Swedhaa, M.A. Sathya, V. Vanitha, M. Jothilakshmi, T. Kongunadu College of Engineering and Technology Artificial Intelligence and Data Science Trichy India Mohamed Sathak A.J. College of Engineering Computer Science and Engineering Tamilnadu Chennai India

ISBN: (纸本)9798350349221

Cloud Storage will be current data research and data management field in terms of security and elimination of repeated data-sets. In simple terms, this current research introduces a strong system called "Cloud-SecureDedupe(C-SD)". to make sure data is safe and cloud storage works well. The system helps with a common problem called data repetition, where the same information is stored more than once. "Cloud-SecureDedupe (SDP)"uses a safe method to solve this issue, making sure your data is reliable in cloud storage. It adds extra protection by using advanced techniques to keep important information secure. Additionally, it smartly identifies and removes any repeated data, not only making your data more secure but also saving space in the storage, so it's used more efficiently. The system has been carefully tested and proven to work well, ensuring that your experience with cloud storage is both safe and smooth. © 2024 IEEE.

关键词： Cloud storage

来源：评论

学校读者我要写书评

暂无评论

Class-Incremental Few-Shot Event Detection 30

Class-Incremental Few-Shot Event Detection

引用

Joint 30th International Conference on Computational Linguistics and 14th International Conference on Language Resources and Evaluation, LREC-COLING 2024

作者： Zhao, Kailin Jin, Xiaolong Bai, Long Guo, Jiafeng Cheng, Xueqi Key Laboratory of Network Data Science and Technology Institute of Computing Technology Chinese Academy of Sciences China School of Computer Science and Technology University of Chinese Academy of Sciences China

ISBN: (纸本)9782493814104

Event detection is one of the fundamental tasks in information extraction and knowledge graph. However, a realistic event detection system often needs to deal with new event classes constantly. These new classes usually have only a few labeled instances as it is time-consuming and labor-intensive to annotate a large number of unlabeled instances. Therefore, this paper proposes a new task, called class-incremental few-shot event detection. Nevertheless, this task faces two problems, i.e., old knowledge forgetting and new class overfitting. To solve these problems, this paper further presents a novel knowledge distillation and prompt learning based method, called Prompt-KD. Specifically, to handle the forgetting problem about old knowledge, Prompt-KD develops an attention based multi-teacher knowledge distillation framework, where the ancestor teacher model pre-trained on base classes is reused in all learning sessions, and the father teacher model derives the current student model via adaptation. On the other hand, in order to cope with the few-shot learning scenario and alleviate the corresponding new class overfitting problem, Prompt-KD is also equipped with a prompt learning mechanism. Extensive experiments on two benchmark datasets, i.e., FewEvent and MAVEN, demonstrate the superior performance of Prompt-KD. © 2024 ELRA Language Resource Association: CC BY-NC 4.0.

关键词： Distillation

来源：评论

学校读者我要写书评

暂无评论

Assessing a Personalized, Hybrid, and Generic Approach for Glucose Prediction in Type 1 Diabetes

Assessing a Personalized, Hybrid, and Generic Approach for G...

引用

2024 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2024

作者： Bosoni, Pietro Jensen, Morten Hasselstrom Bellazzi, Riccardo Cichosz, Simon Lebech University of Pavia Department of Electrical Computer and Biomedical Engineering Pavia Italy Aalborg University Data Science Department of Health Science and Technology Aalborg Denmark Aalborg University Department of Health Science and Technology Aalborg Denmark

ISBN: (纸本)9798350386226

This study investigates the potential of using generic, hybrid, and personalized neural network models for glucose prediction in individuals with Type 1 Diabetes (T1D). data from 194 participants in the Wireless Innovations for Seniors with Diabetes Mellitus (WISDM) study, totaling over 46 million minutes of Continuous Glucose Monitoring (CGM), were used to develop and evaluate the models. A baseline reference model, Last Observation Carried Forward (LOCF), was also included for comparison. Models were trained using data from 70% of the participants and tested on the remaining 30%, with prediction horizons (PH) set at 30 and 60 minutes. At the 30-minute PH, the generic model achieved a Root Mean Square Error (RMSE) of 19.6 mg/dL and a Mean Absolute Relative Difference (MARD) of 9.2%. These results were slightly worse than those of the hybrid and personalized models, which yielded RMSEs of 19.4 mg/dL and 19.5 mg/dL, respectively, and MARDs of 9.6% for both. However, the differences were not statistically significant. At the 60-minute PH, the generic model showed the best performance, with an RMSE of 34.9 mg/dL and MARD of 16.2%. The hybrid and personalized models exhibited slightly higher RMSEs (35.3 mg/dL and 35.5 mg/dL, respectively) and MARDs (17.8% and 18.1%, respectively). These findings suggest that both generic and individualized models can provide satisfactory glucose forecasting results based solely on CGM data. Nonetheless, the potential benefits of individualized approaches deserve further investigation, particularly when substantial training data are available. © 2024 IEEE.

关键词： Prediction models

来源：评论

学校读者我要写书评

暂无评论

Analysis of Prediction of Pneumonia from Chest X-ray images using CNN and Transfer Learning 5

Analysis of Prediction of Pneumonia from Chest X-ray images ...

引用

5th IEEE International Conference for Emerging Technology, INCET 2024

作者： Vinay, K. Shashanka Kodipalli, Ashwini Swetha, P. Kumaraswamy, S. Dept. of Computer Science and Engineering Global Academy of Technology Karnataka Bengaluru India Dept of Artificial Intelligence and Data Science Global Academy of Technology Karnataka Bengaluru India

ISBN: (纸本)9798350361155

A good design for image classification problems is the CNN architecture you presented, which consists of 5 convolution blocks followed by 4 fully connected layers. From the input X-ray images, the convolutional blocks extract pertinent features, and the fully connected layers assist in determining the final classification based on those learned features. You have integrated various approaches to improve the performance of your model. The inputs to each layer are normalized through batch normalization, which can speed up training and enhance generalization. By removing certain neurons at random during training, dynamic dropout helps avoid overfitting. L2 regularization weight decay and learning rate decay are two efficient strategies for preventing overfitting and enhancing the model's capacity to expand to new data. Popular optimization algorithm Adam optimizer effectively neural network training. For binary classification problems like the diagnosis of pneumonia, the loss function for binary Cross-Entropy is the best option. To determine your model's efficacy, you must validate it using benchmark datasets that are available to the general public. You can evaluate your model's effectiveness by comparing its performance to that of current methods by conducting experimental investigations on these datasets. Your model performs well as evidenced by accuracy scores of 90.93%, 89.17% for multi-class classification and binary classification. tasks. Automated methods, such as the one you suggested, might help medical practitioners recognize pneumonia and spot diseased spots in chest X-ray pictures. However, it's crucial to remember that automated systems shouldn't take the place of professional radiologists' and doctors' skills and judgment;rather, they should be used as supportive tools. Medical To ensure accurate diagnosis and suitable patient care, specialists should always review and interpret the system's data. It's also crucial to take into account potential drawback

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Don’t Ignore the Drive of Curiosity: Rethinking Subtleties Between Universality of Commonsense Knowledge and Excellence of Large Language Models

引用

SN computer science 2024年第6期5卷 798页

作者： Wang, Chao Chen, Tao Liu, Jingping School of Future Technology Shanghai University Shanghai China Institute of Artificial Intelligence Shanghai University Shanghai China Shanghai Key Laboratory of Data Science School of Computer Science Fudan University Shanghai China School of Information Science and Engineering East China University of Science and Technology Shanghai China

Commonsense reasoning is one of the abilities necessary for artificial intelligence to be as intelligent as humans. However, how to make AI understand commonsense has been a problem that has plagued artificial intelligence for more than 60 years. Existing efforts focus more on the means of knowledge acquisition and strive to enrich the capacity of commonsense knowledge (CSK) bases and dimensions of CSK through advanced methods. Unfortunately, this exuberance has obscured a general consideration of CSK, such as how to follow human habits to obtain the most representative knowledge we need to understand the world. In this paper, this representative knowledge is referred to as core CSK. The influence of core CSK is extensive, and it constitutes almost the fundamental element of human life and the most fundamental cognition of the world. Harnessing human curiosity to find solutions to the above problems is an effective and straightforward route. Specifically, we focus on a special corpus to mine core CSK, namely, why-questions. For example, we can harvest “the sky is blue” from “why is the sky blue?”. To this end, we propose a novel method to extract CSK from why-questions, which mainly consist of two modules. The first is a question classification module used to determine whether a question contains CSK. In this module, we propose a classifier based on a one-sided bootstrapping method and design several informative features for the classifier. The second is a crowdsourcing module used to improve the quality of the extracted commonsense. We conduct extensive experiments, and the experimental results show that our method effectively mines CSK from question corpora. Furthermore, statistical analysis demonstrates the feasibility of this curiosity-driven approach, implying that we provide a basic idea for collecting core CSK. Remarkably, today’s outstanding large language models do not have such simple knowledge summarization capabilities, demonstrating the barrier between

关键词： Classification Core Commonsense knowledge Crowdsourcing Curiosity Psychologically

来源：评论

学校读者我要写书评

暂无评论

Enhancing data sovereignty to improve intelligent mobility services in smart cities

Urban Governance

引用

Urban Governance 2025年第1期5卷 20-31页

作者： Anthony, Bokolo Sarshar, Sizarta Department of Applied Data Science Institute for Energy Technology Halden 1777 Norway Department of Computer Science and Communication Østfold University College Halden Norway Department of Risk and Safety Institute for Energy Technology Halden 1777 Norway

Smart cities aim to provide more digitalized, equitable, sustainable, and liveable cities. In smart cities data evolves as an important asset and citizens data in particular is being used to provide data-driven mobility services. Likewise, in smart cities data is produced and used by individuals having no ownership or control over these data thus affecting their security and privacy. Also, actors such as mobility service providers, companies, third parties, etc. often leverage individuals’ data without regarding users’ autonomy. Accordingly, data sovereignty which entails the possibility to govern and keep control over owned data is gaining widespread attention. As a result, stakeholders in the mobility sector need to share data whilst keeping control to manage data access and usage. Although, existing initiatives are promoting sovereign participation using a federated infrastructure. However, achieving an effective governance usage control and access mechanisms can be challenging especially in the transportation sector. It is thus imperative for individuals to have control, ownership, and custody over their data. This has necessitated the notions of data sovereignty. Therefore, this article employs a systematic review to examine how data sovereignty can be achieved for individuals when they use urban mobility services in smart cities by enabling the secure sharing of data by giving control of data over to individuals. Additionally, this article designs a data control scheme that can be applied to realize data sovereign mobility services. Evidence from this study provides technical and non-technical requirements needed in realizing data sovereignty in smart cities. © 2025 The Author(s)

关键词： data access mechanisms data governance, data usage control data sovereignty data-driven mobility services Intelligent mobility services Smart cities

来源：评论

学校读者我要写书评

暂无评论

Language-agnostic Zero-shot Entity Linking via Entity Prototype Network 15

Language-agnostic Zero-shot Entity Linking via Entity Protot...

引用

15th IEEE International Conference on Knowledge Graph, ICKG 2024

作者： Yang, Haihong Qin, Ming Zhang, Qiang Xu, Renjun Zhejiang University College of Computer Science and Technology Hangzhou China Zhejiang University School of Software Technology Hangzhou China Zhejiang University Center for Data Science Hangzhou China

ISBN: (纸本)9798331508821

Following the recent trend of language understanding, we introduce a novel approach for constructing a general-purpose entity linking system by learning entity prototypes. Our model, the Entity Prototype Network (EPN), generates language-agnostic entity prototypes by reading entity descriptions and aligning diverse surface forms in different languages to the corresponding entity prototype in a unified vector space. To validate our core assumption of learnable language-agnostic entity prototypes, we define a new Zero-shot Cross-lingual Entity Linking (ZXEL) task. Additionally, we propose a simple yet effective auxiliary task termed Entity Identification, which enhances our model's ability to capture both intra-entity similarity and inter-entity dissimilarity. We evaluate our model on the new task in both classic and generalized settings. Experimental results demonstrate consistent improvements (exceeding 30% on average) over competitive baselines. An ablation study confirms the necessity of our model design and highlights the impact of the proposed auxiliary task. © 2024 IEEE.

关键词： Zero-shot learning

来源：评论

学校读者我要写书评

暂无评论

Provenance model for Lattice QCD 40

Provenance model for Lattice QCD

引用

40th International Symposium on Lattice Field Theory, LATTICE 2023

作者： Auge, Tanja Bali, Gunnar Klettke, Meike Ludäscher, Bertram Söldner, Wolfgang Weishäupl, Simon Wettig, Tilo University of Regensburg Faculty of Computer Science and Data Science Germany University of Regensburg Department of Physics Germany University of Illinois at Urbana-Champaign School of Information Sciences United States

Workflow management has become an important topic in many research communities. Here, we focus on the particular aspect of provenance tracking. We follow the W3C PROV standard and formulate a provenance model for Lattice QCD that includes the ensemble-generation and measurement parts of the Lattice QCD workflow. Since many important provenance questions in our community require extensions of this model, we propose a multi-layered provenance approach that combines prospective and retrospective elements. © Copyright owned by the author(s) under the terms of the Creative Commons.

关键词： Lattice theory

来源：评论

学校读者我要写书评

暂无评论

The impact and prospects of computer-aided technology for promoting universal breast cancer screening 2

The impact and prospects of computer-aided technology for pr...

引用

2nd International Conference on Health Big data and Intelligent Healthcare, ICHIH 2023

作者： Zhu, Tianyuan Fu, Xiaoyang Huang, Yucheng Zhuhai College of Science and Technology Alibaba Cloud Big Data Application School Zhuhai China Zhuhai College of Science and Technology School of Computer Science Zhuhai China

ISBN: (纸本)9798350309324

Since breast cancer is difficult to prevent, detecting and treating it early may improve patient outcomes. Universal breast cancer screening is an important method for early detection and treatment. Our work surveyed different populations about their awareness of breast cancer and their willingness to be screened. Then, we summarized the potential barriers to universal breast cancer screening through a comparison of current breast cancer screening methods and models. We propose using computer-aided technology for promoting universal breast cancer screening, including predicting disease risk, improving diagnostic efficiency, and achieving individualized screening goals. Finally, we analyze the problems that need to be solved, outline the basic steps to implement improvements in the integration of computer-aided technology into the health care field represented by breast cancer screening and describe the prospects for future development. © 2023 IEEE.

关键词： Diseases

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：