检索结果-内蒙古大学图书馆

LULC-SegNet: Enhancing Land Use and Land Cover Semantic Segmentation with Denoising Diffusion Feature Fusion

REMOTE SENSING 2024年第23期16卷 4573-4573页

作者： Shi, Zongwen Fan, Junfu Du, Yujie Zhou, Yuke Zhang, Yi Shandong Univ Technol Sch Civil Engn & Geomat Zibo 255000 Peoples R China Nanjing Univ Informat Sci & Technol Sch Geog Sci Nanjing 210044 Peoples R China Chinese Acad Sci Inst Geog Sci & Nat Resources Res State Key Lab Resources & Environm Informat Syst Beijing 100101 Peoples R China Shandong Agr & Engn Univ Coll Land Resources & Surveying Engn Zibo 255300 Peoples R China Cent South Univ Sch Geosci & Info Phys Changsha 410083 Peoples R China

Deep convolutional networks often encounter information bottlenecks when extracting land object features, resulting in critical geometric information loss, which impedes semantic segmentation capabilities in complex geospatial backgrounds. We developed LULC-SegNet, a semantic segmentation network for land use and land cover (LULC), which integrates features from the denoising diffusion probabilistic model (DDPM). This network enhances the clarity of the edge segmentation, detail resolution, and the visualization and accuracy of the contours by delving into the spatial details of the remote sensing images. The LULC-SegNet incorporates DDPM decoder features into the LULC segmentation task, utilizing machine learning clustering algorithms and spatial attention to extract continuous DDPM semantic features. The network addresses the potential loss of spatial details during feature extraction in convolutional neural network (CNN), and the integration of the DDPM features with the CNN feature extraction network improves the accuracy of the segmentation boundaries of the geographical features. Ablation and comparison experiments conducted on the Circum-Tarim Basin Region LULC Dataset demonstrate that the LULC-SegNet improved the LULC semantic segmentation. The LULC-SegNet excels in multiple key performance indicators compared to existing advanced semantic segmentation methods. Specifically, the network achieved remarkable scores of 80.25% in the mean intersection over union (MIOU) and 93.92% in the F1 score, surpassing current technologies. The LULC-SegNet demonstrated an IOU score of 73.67%, particularly in segmenting the small-sample river class. Our method adapts to the complex geophysical characteristics of remote sensing datasets, enhancing the performance of automatic semantic segmentation tasks for land use and land cover changes and making critical advancements.

关键词： land use and land cover semantic segmentation remote sensing images denoising diffusion probabilistic model feature fusion k-means clustering algorithms

来源：评论

学校读者我要写书评

暂无评论

Automatic cloud segmentation from INSAT-3D satellite image via IkM and IFCM clustering

引用

IET IMAGE PROCESSING 2020年第7期14卷 1273-1280页

作者： Pugazhenthi, A. kumar, Lakshmi Sutha Natl Inst Technol Puducherry Dept Elect & Commun Engn Karaikal Union Territory India

Cloud extraction and classification from satellite imagery is important for many applications in remote sensing. Satellite images are segmented based on distance, intensity and texture of the images. The popular segmentation algorithms, k-means (kM) and fuzzy c-means (FCM) clustering algorithms, face some problems such as unknown number of groups, unknown initialization and dead centers. In this paper, an unsupervised pixel classification by the kM and FCM algorithms is improved and the selection of centroids is made automatic. The proposed improved k-means (IkM) and improved fuzzy c-means (IFCM) clustering algorithms segment the INSAT-3D satellite's thermal infrared image into low-level, middle-level, high-level clouds and non-cloudy region. As human beings can easily find the clouds in the satellite images, visible image is used to differentiate the clouds from the background. A threshold is found from the histogram of the visible image to separate the cloudy and non-cloudy pixels. The other three thresholds to divide the clouds into three types are found from the thermal infrared image's histogram. The segmentation results of IkM and IFCM algorithms are compared with the existing segmentation algorithms. The comparison shows that IFCM algorithm matches well with original image followed by IkM algorithm as compared with existing algorithms.

关键词： clouds image classification fuzzy set theory geophysical image processing image segmentation remote sensing atmospheric techniques automatic cloud segmentation IFCM clustering satellite imagery satellite images image texture FCM clustering algorithms unsupervised pixel classification low-level clouds middle-level clouds high-level clouds noncloudy region visible image VIS image cloudy IkM segmentation algorithm IFCM segmentation algorithm low-level cloud region middle-level cloud region high-level cloudregion satellite system-three-dimensional satellite image k-means clustering algorithms fuzzy c-means clustering algorithms TIR image histogram

来源：评论

学校读者我要写书评

暂无评论

Forecasting Trading-Time based Profit-Making Strategies in Forex Industry: Using Australian Forex Data 11

Forecasting Trading-Time based Profit-Making Strategies in F...

引用

11th International Conference on knowledge and Systems Engineering (kSE)

作者： Rupasinghe, Mihiran Halgamuge, Malka N. Nguyen Tran Quoc Vinh Technium eCommerce Melbourne Vic 3000 Australia Univ Melbourne Dept Elect & Elect Engn Melbourne Vic 3010 Australia Univ Da Nang Univ Sci & Educ Fac Informat Technol Da Nang Vietnam

ISBN: (纸本)9781728130033

Due to the constant fluctuation on global currency rates, it is challenging to make predictions on trading in foreign exchange (Forex) currency market without an intensive analysis;hence, traders struggle to make a profit. This study aims to analyze the relationship between the trade open time and profit in the Forex currency market to help traders to increase the chance of winning trades and make a profit. We developed a technique to observe the most suitable time duration to trade and the profit. This technique assists traders to enhance the chance of winning trades and make a profit by identifying whether it is more likely to make a profit when they keep the trade opened for a longer time or a shorter time. A Forex dataset (N=1,000,000 trades) from a third-party broker database based in Australia has been used. The collected data were filtered according to the popularity of currency pairs. Five currency pairs (as EUR vs USD, GBP vs JPY, USD vs JPY, GBP vs USD and EUR vs JPY) were further analyzed using Support Vector Machine (SVM) with the Radial Basis Function (RBF) kernel and k-means clustering algorithms. It showed that EUR vs USD and USD vs JPY have sensitive movements of profit with the trading time. The highest profit was observed trading time in between 5 to 15 minutes. Our analysis illustrates that shorter time traders are making more profits than the longer time traders. Hence, this study demonstrates that Forex traders make a profit when the market has a unique volatile situation. This study should be useful as a reference for researches in Forex market analyses and Forex Industry to utilize profit-making strategies.

关键词： Forex industry Volatile market Open price Close price Volume Currency rate Support Vector Machine k-means clustering algorithms

来源：评论

学校读者我要写书评

暂无评论

Applying Machine Learning Classification to Determining Outliers in Effort for Embedded Software Development Projects 6

Applying Machine Learning Classification to Determining Outl...

引用

6th ACIS International Conference on Computational Science/Intelligence and Applied Informatics (CSII)

作者： Iwata, kazunori Nakashima, Toyoshiro Anan, Yoshiyuki Ishii, Naohiro Aichi Univ Dept Business Adm Nakamura Ku 4-60-6 Hiraike Cho Nagoya Aichi 4538777 Japan Aichi Univ Inst Managerial Res Nakamura Ku 4-60-6 Hiraike Cho Nagoya Aichi 4538777 Japan Omron Software Co Ltd Informat & Syst Technol Div Proc Engn Dept Nakagyo Ku Karasuma Chuo Bldg 8F659 Tearaimizucho Kyoto 6048152 Japan Aichi Inst Technol Dept Informat Sci 1247 YachigusaYakusa Cho Toyota Aichi 4700392 Japan

ISBN: (纸本)9781728125534

This paper discusses the effect of under-sampling to determine outliers in terms of the amount of effort, which is the number of people hours needed to complete a task, in embedded software development projects. Estimating the amount of effort for new software projects is especially important. As outliers are harmful to the estimation, they are excluded from many estimation models. However, such outliers can be identified in practice once the projects are completed, and so they should not be excluded during the creation of models and when estimating the amount of effort. We have also attempted to detect outliers in the amount of effort by using an ANN and an SVM. However, the accuracy of the classifications was insufficient due to a small number of outliers. This problem is called data imbalance, and occurs in most of machine learning methods. In order to avoid this problem, we explores rebalancing methods with k-means cluster-based under-sampling. The purpose of the methods are to improve the proportion of outliers that are correctly identified, while keeping the other classification criteria high. Evaluation experiments are carried out to compare the prediction accuracy of the methods with k-means under-sampling, with random under-sampling and without under-sampling using 10-fold cross-validation. The results show that the methods with under-sampling have higher sensitivity and lower precision than these of the methods without under-sampling. The results show that the proposed methods can improve the accuracy to determine outliers;however they pick up excess samples as outliers.

关键词： support vector machine artificial neural network embedded software imbalanced data sets k-means clustering algorithms under-sampling

来源：评论

学校读者我要写书评

暂无评论

An Energy Efficient Tour Construction Using Restricted k-means clustering Algorithm for Mobile Sink in Wireless Sensor Networks 11

An Energy Efficient Tour Construction Using Restricted k-mea...

引用

11th International Conference on Developments in eSystems Engineering (DeSE)

作者： Rasul, Aram Al-Talabani, Abdulbasit Univ Raparin Dept Comp Sci Ranya Iraq Koya Univ Dept Software Engn Koya Iraq

ISBN: (纸本)9781538667125

The abstract There has been much research on efficient energy utilisation to prolong the life-span of wireless sensor networks and other tiny devices, with various techniques deployed to address energy consumption issues. The aim of this paper is to build on previous research and further investigate the use of a mobile sink for data collection in wireless sensor networks. We aim to find an optimal path for a mobile sink to collect a single packet from each sensor via a single hop and return back to the starting point such that, subject to the length constraint L, total energy wastage is minimised. We have previously referred to this problem as the minimum energy cost mobile sink restricted tour problem and showed that this is NP-hard. We were inspired by the concept of the k-means clustering algorithm and propose a restricted k-means clustering algorithm. In this approach, we first divide the sensing field into a set of k clusters such that the radius of each cluster is R, where R is the maximum transmission range of the sensor. We iteratively increase the value of k until all the sensors are covered under the length constraint. Simplicity, efficiency, and flexibility are the most important and distinctive features of this algorithm. The technique is implemented to evaluate the algorithm and compare it to our previous algorithm. Our simulation results outperformed the previous technique.

关键词： Wireless sensor networks k-means clustering algorithms Mobility Heuristic algorithms Energy consumption disk covering problem covering salesman problem

来源：评论

学校读者我要写书评

暂无评论

Visual k-means Approach for Handling Class Imbalance Learning 2nd

Visual K-Means Approach for Handling Class Imbalance Learnin...

引用

2nd International Conference on Computer and Communication Technologies

作者： kumar, Ch. N. Santhosh Rao, k. Nageswara Govardhan, A. JNTUH Dept CSE Hyderabad Andhra Pradesh India PSCMR Coll Engn & Technol Vijayawada AP India JNTUH SIT CSE Hyderabad Andhra Pradesh India

ISBN: (纸本)9788132225263;9788132225256

In this paper, a novel clustering algorithm dubbed as Visual k-means (VkM) is proposed. The proposed algorithm deals with the uniform effect which is very much visible in k-means algorithm for skewed distributed data sources. The evaluation of the proposed algorithm is conducted with 10 imbalanced dataset against five benchmark algorithms on six evaluation metrics. The observations from the simulation results project that the proposed algorithm is one of the best alternatives to handle the imbalanced datasets effectively.

关键词： Imbalanced data k-means clustering algorithms Under sampling Visual k-means

来源：评论

学校读者我要写书评

暂无评论

Subset k-means Approach for Handling Imbalanced-Distributed Data 49

Subset K-Means Approach for Handling Imbalanced-Distributed ...

引用

49th Annual Convention of Computer-Society-of-India - Emerging ICT for Bridging the Future

作者： kumar, Ch N. Santhosh Rao, k. Nageswara Govardhan, A. Sandhya, N. JNTU Hyderabad Dept CSE Hyderabad Telangana India PSCMR Coll Engn & Technol Vijayawada AP India JNTU Hyderabad SIT CSE Hyderabad Telangana India VNR Vignana Jyothi Inst Engn & Technol CSE Dept Hyderabad Andhra Pradesh India

ISBN: (纸本)9783319137315;9783319137308

The effectiveness of clustering analysis relies not only on the assumption of cluster number but also on the class distribution of the data employed. This paper represents another step in overcoming a drawback of k-means, its lack of defense against imbalance data distribution. k-means is a partitional clustering technique that is well-known and widely used for its low computational cost. However, the performance of k-means algorithm tends to be affected by skewed data distributions, i.e., imbalanced data. They often produce clusters of relatively uniform sizes, even if input data have varied cluster size, which is called the "uniform effect." In this paper, we analyze the causes of this effect and illustrate that it probably occurs more in the k-means clustering process. As the minority class decreases in size, the "uniform effect" becomes evident. To prevent the effect of the "uniform effect", we revisit the well-known k-means algorithm and provide a general method to properly cluster imbalance distributed data. The proposed algorithm consists of a novel under random subset generation technique implemented by defining number of subsets depending upon the unique properties of the dataset. We conduct experiments using ten UCI datasets from various application domains using five algorithms for comparison on eight evaluation metrics. Experiment results show that our proposed approach has several distinctive advantages over the original k-means and other clustering methods.

关键词： data k-means clustering algorithms oversampling k-Subset

来源：评论

学校读者我要写书评

暂无评论

Undersampled k-means approach for handling imbalanced distributed data

引用

PROGRESS IN ARTIFICIAL INTELLIGENCE 2014年第1期3卷 29-38页

作者： kumar, N. Santhosh Rao, k. Nageswara Govardhan, A. Reddy, k. Sudheer Mahmood, Ali Mirza JNTU Dept CSE Hyderabad Andhra Prades India PSCMR Coll Engn & Technol Vijayawada Andhra Prades India CSE Hyderabad Andhra Prades India JNTU SIT Hyderabad Andhra Prades India Infosys Hyderabad Andhra Prades India DMS SVH Coll Engn Machilipatam Andhra Prades India

k-means is a partitional clustering technique that is well known and widely used for its low computational cost. However, the performance of k-means algorithm tends to be affected by skewed data distributions, i.e., imbalanced data. They often produce clusters of relatively uniform sizes, even if input data have varied cluster size, which is called the "uniform effect". In this paper, we analyze the causes of this effect and illustrate that it probably occurs more in the k-means clustering process. As the minority class decreases in size, the "uniform effect" becomes evident. To prevent the effect of the "uniform effect", we revisit the well-known k-means algorithm and provide a general method to properly cluster imbalance distributed data. The proposed algorithm consists of a novel undersampling technique implemented by intelligently removing noisy and weak instances from majority class. We conduct experiments using twelve UCI datasets from various application domains using five algorithms for comparison on eight evaluation metrics. Experimental results show the effectiveness of the proposed clustering algorithm in clustering balanced and imbalanced data.

关键词： Imbalanced data k-means clustering algorithms Undersampling USkM

来源：评论

学校读者我要写书评

暂无评论

Modelling of Combustion Process Characteristics of a 600 MW Boiler by T-S Fuzzy Neural Networks

Modelling of Combustion Process Characteristics of a 600 MW ...

引用

International Conference on Energy and Environment Technology (ICEET 2009)

作者： Huang Xian Wang Hui N China Elect Power Univ Dept Automat Beijing Peoples R China

ISBN: (纸本)9780769538198

Taking thermal efficiency and LOI as outputs and 20 operating parameters as inputs of the combustion process of a 600 MWe cyclone boiler with 6 mills, 10 fuzzy rules are set up by k-means clustering firstly according to the data collected by DCS. Secondly, a T-S (Takagi-Sugeno) fuzzy neural network which consists of 11 BP sub-networks is developed to model the combustion process characteristics. As the LOI measuring device does not work well and thermal efficiency is a calculation value which is dependent on other primary parameters especially those of coal quality which are hard to be measured accurately and timely, the data collected are unavoidably spoiled by noise seriously. Normal neural networks such as BP and RBF are not capable of dealing with such a noise case. Validation tests demonstrate that the proposed T-S fuzzy neural network, which synthesizes well the advantages both of neural network and fuzzy reasoning, is much friendly to data noise and can be chose as an applied data mining tool.

关键词： boiler combustion k-means clustering algorithms T-S FNN BP ANN optimization

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：