检索结果-内蒙古大学图书馆

20th International Conference on Smart Technologies, EUROCON 2023

作者： Salii, Yevhenii Kuzin, Volodymyr Hohol, Anton Kussul, Nataliia Yailymova, Hanna National Technical University of Ukraine 'Igor Sikorsky Kyiv Polytechnic Institute' Department of Mathematical Modeling and Data Analysis Kyiv Ukraine National University of 'Kyiv-Mohyla Academy' Department of Informatics Kyiv Ukraine National Technical University of Ukraine 'Igor Sikorsky Kyiv Polytechnic Institute' Department of Space Information Technologies and Systems Space Research Institute NASU-SSAU Department of Mathematical Modeling and Data Analysis Kyiv Ukraine

ISBN: (纸本)9781665463973

The paper deals with the problem of semantic segmentation of satellite imagery to deliver forest type map with high resolution. To solve the problem, we propose 4 machine learning models. Two of them are based on Random Forest (RF) and other two - on Convolutional Neural Network (CNN) - U-Net. As an input we use 2 images of Sentinel-2 (one for summer and one for winter, 4 spectral bands from each). As an output (labels) we use the Copernicus Forest Type dataset for 2018 year. Our models showed promising results on validation data. Of all models the one based on U-Net ended up being the most efficient in forest classification with overall accuracy 91.7%. At the same time the best RF model scored with 86.5%. After comparing the results, in order to check our model transferability we created and compared forest map of northern part of Kyiv region of 2018 and 2022. The experiment confirmed the robustness of the model and it's scalability. The developed models have been implemented in the cloud platform specialized on satellite data - CREODIAS. The developed map can provide valuable data for foresters, biologists, or other researchers to make decisions about forest management and conservation, as well as to ensure that Europe's forests are managed in an ecologically sustainable way. © 2023 IEEE.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

The Optimal Closed-Form Linear Canonical Wigner Distribution 6

The Optimal Closed-Form Linear Canonical Wigner Distribution

引用

6th International Conference on Information Communication and Signal Processing, ICICSP 2023

作者： Zhang, Zhichao School of Mathematics and Statistics Center for Applied Mathematics of Jiangsu Province Nanjing University of Information Science and Technology Jiangsu International Joint Laboratory on System Modeling and Data Analysis Nanjing China

ISBN: (纸本)9798350339994

This study devotes to Heisenberg's uncertainty inequalities under the closed-form instantaneous cross-correlation function Wigner distribution (CICFWD), whose generalizability enables it to be a representative Wigner distribution (WD) in the linear canonical transform (LCT) domain. A relation between the spread in the time-CICFWD domain and the spreads and moments in LCT domains is established. We then obtain two versions of lower bounds on the uncertainty product of spreads in time-CICFWD and LCT-CICFWD domains. The first is the tightest universal lower bound for all signals, and the other is a lower bound for complex-valued signals. As a result, uncertainty principles for the traditional WD are revisited. The optimal LCT free parameters selection problems of the CICFWD are also well-addressed from the perspectives of the highest time-frequency resolution and its best estimation accuracy, respectively. © 2023 IEEE.

关键词： linear canonical Wigner distribution lower bound time-frequency resolution uncertainty principle

来源：评论

学校读者我要写书评

暂无评论

Referable Diabetic Retinopathy Detection Using Deep Feature Extraction and Random Forest 15th

Referable Diabetic Retinopathy Detection Using Deep Feature ...

引用

Proceedings of the 15th International Conference on Biomedical Engineering Systems and Technologies, BIOSTEC 2022

作者： Lahmar, Chaymaa Idri, Ali Software Project Management Research Team ENSIAS Mohammed V University in Rabat Rabat Morocco Modeling Simulation and Data Analysis Mohammed VI Polytechnic University Benguerir Ben Guerir Morocco

ISBN: (纸本)9783031388538

Diabetic retinopathy (DR) is the most common eyes complication of diabetes worldwide;it can cause vision loss and blindness. The early diagnosis can significantly help in assuring an effective treatment. The computer vision techniques are playing an important role in improving the diagnosis results. This paper proposes a hybrid architecture that combines: four of the most recent deep learning techniques for feature extraction (DenseNet_201, MobileNet_V2, VGG16 and VGG19) with a random forest classifier for referable diabetic retinopathy detection over the APTOS, Kaggle DR and Messidor-2 datasets. The study evaluated and compared: (1) the random forest models with their base learners, (2) the designed random forest models with the same feature extractor over different number of trees, (3) the decision tree classifiers with the best random forest models and (4) the best random forest models of the four feature extractors to each other. The empirical evaluations used: four classification performance criteria (accuracy, sensitivity, precision and F1-score), 5-fold cross-validation, Scott Knott statistical test, and Borda Count voting method. The best model was constructed using a random forest classifier of 9 trees with MobileNet_V2 for feature extraction, it was trained over the APTOS dataset, and it achieved an accuracy value of 82.12%. The experimental results demonstrated that combining random forest with deep learning models is effective for referable diabetic retinopathy detection using fundus images. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Robust Low-Rank Matrix Completion Under Generalized Gaussian Mixture Model With Application to DoA Estimation

Robust Low-Rank Matrix Completion Under Generalized Gaussian...

引用

IET International Radar Conference 2023, IRC 2023

作者： Tang, Mengjiao Zan, Zeqiong Rong, Yao Hu, Sanfeng Li, Fan Yunnan Key Laboratory of Statistical Modeling and Data Analysis Yunnan University Kunming China CAAC Key Laboratory of Flight Techniques and Flight Safety Civil Aviation Flight University of China Guanghan China

ISBN: (纸本)9781839539954

This paper considers the problem of noisy low-rank matrix completion (LRMC), presenting a novel and robust LRMC approach that differs from classical methods reliant solely on Lp-norms for handling single noise distributions. Instead, we adopt a model that accounts for the noise using a Mixture of Generalized Gaussian (MoGG) distributions'in both the real and complex circularly symmetric cases'thus accommodating more intricate noise patterns. Our proposed method formulates the LRMC problem as the minimization of the negative log-likelihood coupled with a nuclear norm penalty, which is effectively addressed by the EM-ADMM algorithm devised in this paper. Moreover, the proposed complex-valued LRMC method is applied to tackle the Direction of Arrival (DoA) estimation with incomplete measurements. Specifically, we leverage the developed LRMC technique for imputation of the missing entries and then employ the MUSIC algorithm to estimate the angles of impinging signals. Finally, numerical examples are presented to show the performance improvement of our proposed method compared to some classical ones. © The Institution of Engineering & Technology 2023.

关键词： COMPLEX CIRCULARLY SYMMETRIC DISTRIBUTIONS DOA ESTIMATION LOW-RANK MATRIX COMPLETION MIXTURE OF GENERALIZED GAUSSIAN DISTRIBUTIONS

来源：评论

学校读者我要写书评

暂无评论

Algorithms of Time Series Network: Approaches Reproduction and Networks Topology 9

Algorithms of Time Series Network: Approaches Reproduction a...

引用

9th International Conference on Fuzzy Systems and data Mining, FSDM 2023

作者： Wang, Li-Na Cheng, Yuan-Yuan Tan, Gui-Min Science College Inner Mongolia University of Technology Hohhot010051 China Inner Mongolian Key Lab. of Statistical Analysis Theory for Life Data and Neural Network Modeling Hohhot010051 China

ISBN: (纸本)9781643684703

Network analysis methods of time series provide a new analytical framework on describing complex behaviors using sample data. Reproducibility is one of the important principles in scientific research. This work focuses on the algorithm review and reproduction of time series network. We paid special attention to the applicability and data characteristics of the five typical algorithms: visibility graph, phase space reconstruction, fluctuation mode, symbolic representation and coarsened multidimensional time series. It provides a reference and inspiration for future analysis of time series with various characteristics. Although two pioneer methods are widely applicable, while the directed weighted network established by coarsening thoughts contains more information about time series. In addition, coarseness process makes these approaches perform better for massive data analysis. It should be noted that: the process of data coarsening needs to ensure that the data characteristics of the original time series are inherited. © 2023 The authors and IOS Press.

关键词： Time series

来源：评论

学校读者我要写书评

暂无评论

Can we promote neural regeneration through microbiota-targeted strategies? Introducing the new concept of neurobiotics

引用

Neural Regeneration Research 2022年第9期17卷 1965-1966页

作者： Celia Herrera-Rincon Julia Murciano-Brea Stefano Geuna Department of Biodiversity Ecology&Evolutionand ModelingData Analysis&Computational Tools for Biology Research GroupBiomathematics UnitComplutense University of MadridMadridSpain Department of Clinical and Biological Sciences and Cavalieri Ottolenghi Neuroscience InstituteUniversity of TurinTurinItaly

The human body is populated by a large number of microbial colonies,with an estimated 10-100 trillion *** total genome size of human microbial colonies by far overwhelms the size of the host’s *** heterogenous group of microbial colonies(primarily bacteria,but also archaea,eukaryotes and viruses)is referred to with the term microbiota,and although most of them populate the gut,microbes are also detectable in many other organs of the body,especially in the distal tracts of the genitourinary system and the *** the last years,an increasing amount of evidence has been accumulated on how the microbiota exerts a significant influence on the development and physiology of the human body.

关键词： neural organs size

来源：评论

学校读者我要写书评

暂无评论

Assessing the Impact of Deep End-to-End Architectures in Ensemble Learning for Histopathological Breast Cancer Classification 14

Assessing the Impact of Deep End-to-End Architectures in Ens...

引用

14th International Conference on Knowledge Discovery and Information Retrieval, KDIR 2022 as part of 14th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management, IC3K 2022

作者： Zerouaoui, Hasnae Idri, Ali El Alaoui, Omar Modeling Simulation and Data Analysis Mohammed VI Polytechnic University Benguerir Morocco Software Project Management Research Team ENSIAS Mohammed V University Rabat Morocco

ISBN: (纸本)9789897586149

One of the most significant public health issues in the world and a major factor in women's mortality is breast cancer (BC). Early diagnosis and detection can significantly improve the likelihood of survival. Therefore, this study suggests a deep end-to-end heterogeneous ensemble approach by using deep learning (DL) models for breast histological images classification tested on the BreakHis dataset. The proposed approach showed a significant increase of performances compared to their base learners. Thus, seven DL architectures (VGG16, VGG19, ResNet50, Inception_V3, Inception_ResNet_V2, Xception, and MobileNet) were trained using 5-fold cross-validation. Thereafter, deep end-to-end heterogeneous ensembles of two up to seven base learners were constructed based on accuracy using majority and weighted voting. Results showed the effectiveness of deep end-to-end ensemble learning techniques for breast cancer images classification into malignant or benign. The ensembles designed with weighted voting method exceeded the others with an accuracy value reaching 93.8%, 93.4%, 93.3%, and 91.8% through the BreakHis dataset's four magnification factors: 40X, 100X, 200X, and 400X respectively. Copyright © 2022 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved.

关键词： Learning systems

来源：评论

学校读者我要写书评

暂无评论

Rao Distance Based Spatial-Spectral Joint Feature Extraction and Semi-Supervised Classification for Hyperspectral Image 5

Rao Distance Based Spatial-Spectral Joint Feature Extraction...

引用

5th International Conference on Geoscience and Remote Sensing Mapping, GRSM 2023

作者： Ni, Meixin Tang, Mengjiao Hu, Sanfeng Wang, Liyuqi Li, Fan Yunnan University Yunnan Key Laboratory of Statistical Modeling and Data Analysis Yunnan Kunming650504 China Civil Aviation Flight University of China CAAC Key Laboratory of Flight Techniques and Flight Safety Sichuan Guanghan618307 China

ISBN: (纸本)9798350329551

This paper presents a novel feature extraction approach for hyperspectral images, termed Spatial-Spectral Joint Principal Component analysis (PCA) coupled with Rao distance-based t-distributed Stochastic Neighbor Embedding (SSPCA-Rao-tSNE). Specifically, we apply the classical PCA method for spectral feature extraction while concurrently developing a Rao-distance-based tSNE algorithm to enhance the extraction of spatial features, particularly those with non-Euclidean structural characteristics. These spatial and spectral features are then combined through feature stacking to create spatial-spectral joint features. Furthermore, these spatial-spectral joint features are utilized in a semi-supervised graph convolutional network for hyperspectral image classification. Experimental results demonstrate the superior performance achieved by our newly proposed SSPCA-Rao-tSNE approach and the corresponding semi-supervised classification method in comparison to several classical methods. © 2023 IEEE.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

ON SECOND-ORDER SCORING RULES FOR EPISTEMIC UNCERTAINTY QUANTIFICATION

arXiv

引用

arXiv 2023年

作者： Bengs, Viktor Hüllermeier, Eyke Waegeman, Willem Munich Center for Machine Learning Germany Department of Data Analysis and Mathematical Modeling Ghent University Belgium

It is well known that accurate probabilistic predictors can be trained through empirical risk minimisation with proper scoring rules as loss functions. While such learners capture so-called aleatoric uncertainty of predictions, various machine learning methods have recently been developed with the goal to let the learner also represent its epistemic uncertainty, i.e., the uncertainty caused by a lack of knowledge and data. An emerging branch of the literature proposes the use of a second-order learner that provides predictions in terms of distributions on probability distributions. However, recent work has revealed serious theoretical shortcomings for second-order predictors based on loss minimisation. In this paper, we generalise these findings and prove a more fundamental result: There seems to be no loss function that provides an incentive for a second-order learner to faithfully represent its epistemic uncertainty in the same manner as proper scoring rules do for standard (first-order) learners. As a main mathematical tool to prove this result, we introduce the generalised notion of second-order scoring *** Codes 68T37 (Primary) 68T30 (Secondary) Copyright © 2023, The Authors. All rights reserved.

关键词： Probability distributions

来源：评论

学校读者我要写书评

暂无评论

Accelerating Hybrid Agent-Based Models and Fuzzy Cognitive Maps: How to Combine Agents Who Think Alike? 24

Accelerating Hybrid Agent-Based Models and Fuzzy Cognitive M...

引用

Proceedings of the Winter Simulation Conference

作者： Philippe J. Giabbanelli Jack T. Beerman Virginia Modeling Analysis and Simulation Center Old Dominion University Norfolk VA USA School of Data Science University of Virginia Charlottesville VA USA

ISBN: (纸本)9798331534202

While Agent-Based Models can create detailed artificial societies based on individual differences and local context, they can be computationally intensive. Modelers may offset these costs through a parsimonious use of the model, for example by using smaller population sizes (which limits analyses in sub-populations), running fewer what-if scenarios, or accepting more uncertainty by performing fewer simulations. Alternatively, researchers may accelerate simulations via hardware solutions (e.g., GPU parallelism) or approximation approaches that operate a tradeoff between accuracy and compute time. In this paper, we present an approximation that combines agents who `think alike', thus reducing the population size and the compute time. Our innovation relies on representing agent behaviors as networks of rules (Fuzzy Cognitive Maps) and empirically evaluating different measures of distance between these networks. Then, we form groups of think-alike agents via community detection and simplify them to a representative agent. Case studies show that our simplifications remain accuracy.

关键词：

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：