检索结果-内蒙古大学图书馆

ICT Conference (AICTC), Arab

作者： David Wijaya Ibrahim Seto Aditama Stanley Armando Austen Calvin Surya Widjaja Bakti Amirul Jabar Edy Irwansyah Computer Science Department School of Computer Science Bina Nusantara University Jakarta Indonesia Computer Science Department School of Computer Science Bina Nusantara University Tangerang Selatan Indonesia Computer Science Department School of Computer Science Bina Nusantara University Bandung Indonesia BINUS Graduate Program Master of Computer Science Bina Nusantara University Jakarta Indonesia

ISBN: (数字)9798350342086

ISBN: (纸本)9798350342093

Trees are one of the most important living things on the planet. Trees help by being one of the largest producers' oxygens on planet, by absorbing groundwater and by maintaining soil fertility. Trees can also be one of the keys to reducing existing air pollution. Therefore, the number of existing trees must be maintained. The first step in maintaining the number of existing trees is to count or map them. One of the easiest ways is to do this use an algorithm. This study uses a tree-counting algorithm called Deep Forest. Deep Forest is a Python package that can detect trees through RGB satellite imagery. This study used satellite imagery from Pleiades on an area in Kulon Progo district, Yogyakarta, Indonesia for the image-one of the four regencies within Yogyakarta Province in Indonesia. In addition to the Deep Forest algorithm, this research also used an application provided by Esri called ArcGIS Pro. With this application, data preprocessing such as labeling the data and creating the deep learning dataset for research is much easier, while ArcGIS Pro provides tools for labeling and exporting the data to jupyter notebook. The result obtained using the Deep Forest algorithm is the F1 score with an accuracy of 0.76 for the first experiment, 0.774 for the second experiment, and 0.779 for the final experiment.

关键词： Deep learning Training Codes Planets Vegetation Soil Regression analysis Satellite images Labeling Random forests

来源：评论

学校读者我要写书评

暂无评论

A Stable Intelligent Model-Based Framework Using Graph Convolutional Networks and MULTIMOORA for Clinical Decision Support Systems

引用

IEEE Transactions on Consumer Electronics 2024年第1期71卷 773-784页

作者： Zhang, Chao Li, Xiaochuan Bai, Wenhui Sangaiah, Arun Kumar Li, Deyu Shanxi University School of Computer and Information Technology Key Laboratory of Computational Intelligence and Chinese Information Processing Ministry of Education Taiyuan030006 China National Yunlin University of Science and Technology International Graduate Institute of Artificial Intelligence Douliou64002 Taiwan Sunway University School of Engineering and Technology Selangor Petaling Jaya47500 Malaysia Chandigarh University University Centre for Research and Development Punjab Mohali140413 India

Artificial intelligence (AI) is rapidly advancing in consumer electronics, particularly in clinical decision support systems (CDSSs). AI-driven devices enhance the effectiveness of healthcare-related electronics via predictive health analytics, reshaping medical data collection, analysis, and decision-making. Nevertheless, due to the abundance of imprecise, hesitant, and fuzzy information in healthcare data stemming from consumer electronics, AI may misinterpret the data, resulting in erroneous and unstable medical decisions. Therefore, this paper endeavors to establish a stable intelligent medical decision model, employing two essential tools: graph convolutional networks (GCNs) and MULTIMOORA (Multi-Objective Optimization by Ratio Analysis plus the full MULTIplicative form). First, a granular computing (GrC) framework based on GCNs and MULTIMOORA is established. In this framework, GCNs are initially employed to process multi-modal data within the hesitant fuzzy linguistic (HFL) background. Second, a comprehensive HFL information system (IS) is constructed. Third, three types of multi-granularity (MG) HFL methods are developed via MULTIMOORA. To better address the bounded rationality of decision-makers (DMs), regret theory (RT) is utilized to consolidate the multi-modal data. Finally, the efficacy and practically of the proposed decision model are assessed via a case study on medication selections. © 1975-2011 IEEE.

关键词： Electronic health record

来源：评论

学校读者我要写书评

暂无评论

Multifunctional Ir–Ru alloy catalysts for reversal-tolerant anodes of polymer electrolyte membrane fuel cells

引用

Journal of Materials science & technology 2021年第1期60卷 105-112页

作者： Seung Woo Lee Bongho Lee Chaekyung Baik Tae-Yang Kim Chanho Pak Graduate Program of Energy Technology School of Integrated TechnologyInstitute of Integrated TechnologyGwangju Institute of Science and TechnologyGwangju 61005Republic of Korea Research Institute for Solar and Sustainable Energies Gwangju Institute of Science and TechnologyGwangju 61005Republic of Korea

To address the problem of fuel starvation in fuel-cell electric vehicles,which causes cell voltage reversal and results in cell failure when repeated continuously,we developed a reversal-tolerant anode(RTA) to promote water oxidation in preference to carbon *** carbon-supported Ir-Ru alloys with different compositions are employed as RTA catalysts in an acidic polyol solution and are shown to exhibit composition-dependent average crystallite sizes of <5.33 *** adopted approach allows the generation of relatively well-dispersed Ir-Ru alloy nanoparticles on the carbon support without severe *** activity of IrRu_(2)/C for the hydrogen oxidation reaction is 1.10 times that of the stateof-the-art Pt/C *** reversal testing by simulation of fuel starvation reveals that the durability of IrRu_(2)/C(~7 h) significantly exceeds that of the conventional Pt/C catalyst(~10 min) and is the highest value reported so ***,the developed Ir-Ru alloy catalyst can be used to fabricate practical RTAs and replace Pt catalysts in the anodes of polymer electrolyte membrane fuel cells.

关键词： Polymer electrolyte membrane fuel cell Polyol process Reversal-tolerant anode Oxygen evolution reaction Hydrogen oxidation reaction

来源：评论

学校读者我要写书评

暂无评论

Load-balancing system for SDN data center based on improved ant colony optimization

Load-balancing system for SDN data center based on improved ...

引用

2024 International Conference on computer Network Security and Software Engineering, CNSSE 2024

作者： Zhou, Yue Wu, Wenlong Information Engineering Department Anhui Industry Polytechnic Anhui Tongling244000 China Graduate School La Consolacion University Philippines Bulacan Malolos3000 Philippines Computer Science College Guangzhou College of Applied Science and Technology Guangdong Zhaoqing526072 China

ISBN: (纸本)9781510680234

The load balancing strategy in the traditional network architecture system is limited to the limitations of the closed operating system of the network device and cannot effectively meet various communication environments. In SDN, a new network architecture with open programming, the load balancing system architecture of the SDN data center is designed. The traffic classification module is used to distinguish large and small flows. A load balancing mechanism based on the improved ant colony optimization is proposed for the forwarding of large flows. A weighted ECMP load balancing mechanism is used for the forwarding of small flows. The improved ant colony optimization considers the difference in link load status and optimizes the parameters of the ant colony optimization. Through experimental simulation and comparison with traditional load balancing algorithms ECMP and DLB algorithms, the advantages of the optimized ant colony optimization in terms of delay jitter, average link utilization and other indicators are verified. © 2024 SPIE.

关键词： Ant colony optimization

来源：评论

学校读者我要写书评

暂无评论

Multi-modal Multitask Learning Model for Simultaneous Classification of Two Epilepsy Biomarkers 27th

Multi-modal Multitask Learning Model for Simultaneous Classi...

引用

27th International Conference on Pattern Recognition, ICPR 2024

作者： Broti, Nawara Mahmood Sawada, Masaki Takayama, Yutaro Iijima, Keiya Iwasaki, Masaki Ono, Yumie Electrical Engineering Program Graduate School of Science and Technology Meiji University Kanagawa Kawasaki Japan Department of Neurosurgery Yokohama City University Hospital Kanagawa Yokohama Japan Department of Neurosurgery National Center Hospital National Center of Neurology and Psychiatry Tokyo Kodaira Japan Department of Electronics and Bioinformatics School of Science and Technology Meiji University Kanagawa Kawasaki Japan

ISBN: (纸本)9783031781971

Accurate estimation of the epileptogenic zone is critical for surgical treatment of drug-resistant epilepsy. While epilepsy biomarkers detection for epileptogenic zone localization has traditionally relied on expert clinician knowledge, the use of deep learning in this context has recently gained appeal for objective diagnosis and reducing clinician burden. However, previously proposed classifiers for electrocorticogram data focused on only single biomarkers among many types, requiring separate models and large datasets for each. To minimize clinicians’ workload and patients’ sufferings, innovative methods to classify multiple biomarkers using small training datasets and minimal computation time is required. This research marks the first implementation of multi-modal multitask learning in epilepsy biomarker classification, proposing a model to automatically classify two biomarkers—high-frequency oscillations (HFO) and interictal epileptiform discharges (IED)—from separate datasets simultaneously. We validated the proposed model on 1500 annotated HFO and IED candidate signals each, obtained from 5 patients. Our proposed model can perform both HFO and IED classification simultaneously, with cross-validation accuracies of 90.99% and 93.99%;F1 scores were reported as 0.80 and 0.82, respectively. Our leave-one-patient-out experiments achieved 78.03% and 81.19% accuracy for HFO and IED classification, respectively. Compared to existing state-of-the-art classification architectures, our model is shown to be efficient and robust, yet simple and lightweight. The usefulness of sharing parameters between the two classifiers was confirmed in this practical clinical setting of epileptic biomarker detection, potentially contributing to minimizing the number of annotated datasets and realizing robust, accurate, and simultaneous identification of two biomarkers. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Spiking Transformer with Experts Mixture 38

Spiking Transformer with Experts Mixture

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Zhou, Zhaokun Lu, Yijie Jia, Yanhao Che, Kaiwei Niu, Jun Huang, Liwei Shi, Xinyu Zhu, Yuesheng Li, Guoqi Yu, Zhaofei Yuan, Li School of Electronic and Computer Engineering Shenzhen Graduate School Peking University China Peng Cheng Laboratory China College of Computing and Data Science Nanyang Technological University Singapore School of Computer Science Peking University China Institute for Artificial Intelligence Peking University China Institute of Automation Chinese Academy of Sciences China Deep NeuroCognition Lab I2R and CFAR Agency for Science Technology and Research Singapore

Spiking Neural Networks (SNNs) provide a sparse spike-driven mechanism which is believed to be critical for energy-efficient deep learning. Mixture-of-Experts (MoE), on the other side, aligns with the brain mechanism of distributed and sparse processing, resulting in an efficient way of enhancing model capacity and conditional computation. In this work, we consider how to incorporate SNNs' spike-driven and MoE's conditional computation into a unified framework. However, MoE uses softmax to get the dense conditional weights for each expert and TopK to hard-sparsify the network, which does not fit the properties of SNNs. To address this issue, we reformulate MoE in SNNs and introduce the Spiking Experts Mixture Mechanism (SEMM) from the perspective of sparse spiking activation. Both the experts and the router output spiking sequences, and their element-wise operation makes SEMM computation spike-driven and dynamic sparse-conditional. By developing SEMM into Spiking Transformer, the Experts Mixture Spiking Attention (EMSA) and the Experts Mixture Spiking Perceptron (EMSP) are proposed, which performs routing allocation for head-wise and channel-wise spiking experts, respectively. Experiments show that SEMM realizes sparse conditional computation and obtains a stable improvement on neuromorphic and static datasets with approximate computational overhead based on the Spiking Transformer baselines. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Microservices Architecture As a Data Governance Tools in Decentralized E-Government

Microservices Architecture As a Data Governance Tools in Dec...

引用

2022 International Conference on science and technology, ICOSTECH 2022

作者： Paramita, Adi Suryaputra Gaol, Ford Lomban Ranti, Benny Supangkat, Suhono Harso Graduate Program Universitas Ciputra Surabaya/ Bina Nusantara University Information Systems Department/Doctor of Computer Science Jakarta Surabaya Indonesia Bina Nusantara University Doctor of Computer Science Graduate Program Jakarta Indonesia Institute of Technology Bandung Systems and Information Technology Department Bandung Indonesia

ISBN: (纸本)9781665479851

Electronic government, or e-Government, is the use of information and communication technology by the public sector (ICT). Additionally, it can be viewed as a paradigm shift in terms of how governments operate. The government protects public values, while citizens act as problem solvers, co-creators, and active participants in value creation. On the other hand, decentralized systems have grown in popularity over the last few years. Data has become a significant issue in decentralized E-Government;each local entity has their own data, which is private. A suitable data governance tool is required to ensure that data governance is properly implemented and to prevent the organizational entity from leaking information. The purpose of this research is to examine the data governance tools mechanism utilizing microservices. © 2022 IEEE.

关键词： Information use

来源：评论

学校读者我要写书评

暂无评论

BERT-based Topic Modeling Approach for Malaria Research Publication 4

BERT-based Topic Modeling Approach for Malaria Research Publ...

引用

4th International Conference on Informatics, Multimedia, Cyber and Information System, ICIMCIS 2022

作者： Hidayat, Alam Ahmad Nirwantono, Rudi Budiarto, Arif Pardamean, Bens Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta11480 Indonesia School of Computer Science Bina Nusantara University Computer Science Department Jakarta11480 Indonesia Bina Nusantara University Binus Graduate Program - Master of Computer Science Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9781665473279

Malaria is a communicable disease with half of the global population at risk due to its high morbidity and mortality rates. A massive number of studies are dedicated to malaria research, so it plays a key role in formulating the proper prevention strategy and effective malaria treatment. With the overwhelming number of updated publications in the field, an unsupervised text mining approach such as topic modeling may provide an alternative method for the malaria researcher to keep pace with new insights. In this work, we collect metadata of malaria publications from the PubMed database to perform BERT-based topic modeling to find well-defined topics regarding malaria research. The method is largely based on the popular BERTopic pipeline. We compare the performance of three different language models to generate document embeddings from the data. The dimension reduction and the density-based clustering algorithm are used to cluster the embeddings. The topic representation is computed based on the semantic similarity of the class TF-IDF representation. The substance of the resulting topics is then manually annotated based on the top words of each topic. We demonstrate that by merging initial topics into larger topics using hierarchical clustering and manual content-based examination, the evaluated coherence measure can be further improved, thus enhancing the topic's interpretability. Our modeling result is able to extract ten major topics recurring in the malaria research publication published from 2017-2022. The result provides preliminary insight to understand the dynamics and patterns of malaria research over the years © 2022 IEEE.

关键词： Diseases

来源：评论

学校读者我要写书评

暂无评论

Fundamental Problems on Bounded-Treewidth Graphs: The Real Source of Hardness 51

Fundamental Problems on Bounded-Treewidth Graphs: The Real S...

引用

51st International Colloquium on Automata, Languages, and programming, ICALP 2024

作者： Esmer, Barış Can Focke, Jacob Marx, Dániel Rzążewski, Pawel CISPA Helmholtz Center for Information Security Saarbrücken Germany Saarbrücken Graduate School of Computer Science Saarland Informatics Campus Germany Warsaw University of Technology Poland University of Warsaw Poland

ISBN: (纸本)9783959773225

It is known for many algorithmic problems that if a tree decomposition of width t is given in the input, then the problem can be solved with exponential dependence on t. A line of research initiated by Lokshtanov, Marx, and Saurabh [SODA 2011] produced lower bounds showing that in many cases known algorithms already achieve the best possible exponential dependence on t, assuming the Strong Exponential-Time Hypothesis (SETH). The main message of this paper is showing that the same lower bounds can already be obtained in a much more restricted setting: informally, a graph consisting of a block of t vertices connected to components of constant size already has the same hardness as a general tree decomposition of width t. Formally, a (σ, δ)-hub is a set Q of vertices such that every component of Q has size at most σ and is adjacent to at most δ vertices of Q. We explore if the known tight lower bounds parameterized by the width of the given tree decomposition remain valid if we parameterize by the size of the given hub. For every Ε > 0, there are σ, δ > 0 such that Independent Set (equivalently Vertex Cover) cannot be solved in time (2 − Ε)p · n, even if a (σ, δ)-hub of size p is given in the input, assuming the SETH. This matches the earlier tight lower bounds parameterized by width of the tree decomposition. Similar tight bounds are obtained for Odd Cycle Transversal, Max Cut, q-Coloring, and edge/vertex deletions versions of q-Coloring. For every Ε > 0, there are σ, δ > 0 such that Δ-Partition cannot be solved in time (2 − Ε)p · n, even if a (σ, δ)-hub of size p is given in the input, assuming the Set Cover Conjecture (SCC). In fact, we prove that this statement is equivalent to the SCC, thus it is unlikely that this could be proved assuming the SETH. For Dominating Set, we can prove a non-tight lower bound ruling out (2 − Ε)p · nO(1) algorithms, assuming either the SETH or the SCC, but this does not match the 3p · nO(1) upper bound. Thus our results reveal that, for

关键词： Hardness

来源：评论

学校读者我要写书评

暂无评论

Deep Image Restoration Model: A Defense Method Against Adversarial Attacks

引用

computers, Materials & Continua 2022年第5期71卷 2209-2224页

作者： Kazim Ali Adnan N.Quershi Ahmad Alauddin Bin Arifin Muhammad Shahid Bhatti Abid Sohail Rohail Hassan Department of Information Technology University of Central PunjabLahore54000Pakistan Department of Communication Technology and Network Faculty of Computer Science and Information TechnologyUniversity Putra MalaysiaSalengor43400Malaysia Department of Computer Science Comsats University IslamabadLahore Campus54000Pakistan Othman Yeop Abdullah Graduate School of Business University Utara MalaysiaKuala Lumpur50300Malaysia

These days,deep learning and computer vision are much-growing fields in this modern world of information *** learning algorithms and computer vision have achieved great success in different applications like image classification,speech recognition,self-driving vehicles,disease diagnostics,and many *** success in various applications,it is found that these learning algorithms face severe threats due to adversarial *** examples are inputs like images in the computer vision field,which are intentionally slightly changed or *** changes are humanly *** are misclassified by a model with high probability and severely affects the performance or *** this scenario,we present a deep image restoration model that restores adversarial examples so that the target model is classified correctly *** proved that our defense method against adversarial attacks based on a deep image restoration model is simple and state-of-the-art by providing strong experimental results *** have used MNIST and CIFAR10 datasets for experiments and analysis of our defense *** the end,we have compared our method to other state-ofthe-art defense methods and proved that our results are better than other rival methods.

关键词： computer vision deep learning convolutional neural networks adversarial examples adversarial attacks adversarial defenses

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：