检索结果-内蒙古大学图书馆

2023 International Conference on Next-Generation Computing, IoT and Machine Learning, NCIM 2023

作者： Pall, Shuvo Chandra Cheragee, Sheikh Hasib Mahamud, Md Raihan Mishu, Hasan Mahmud Biswas, Haraprosad Electrical and Electronic Engineering Bangladesh University of Engineering and Technology Dhaka Bangladesh Institute of Information and Communication Technology Bangladesh University of Engineering and Technology Dhaka Bangladesh Computer Science and Engineering Royal University of Dhaka Dhaka Bangladesh Computer Science and Engineering Bangladesh University of Engineering and Technology Dhaka Bangladesh

ISBN: (纸本)9798350316001

electrical load forecasting is essential for energy suppliers and policymakers to operate and develop power plants and provide stable and dependable energy infrastructure in Bangladesh. Forecasting loads come in three different forms: short-, mid-, and long-term. Of these, forecasting short-term load keeps track of activities on a daily, weekly, and hourly basis, which is becoming more popular since it reduces time and costs. While designing and implementing the electric power system, mid-term forecasting of the load is a critical stage in organizing, maintaining, and preparing for interruptions. On the basis of the Pearson correlation matrix, Machine learning models like support vector regression(SVR) and linear regression(LR), and deep learning, i.e., multilayer perceptron (MLP) approaches presented in this paper. The recommended approaches are evaluated and compared depending on accuracy, mean absolute error (MAE), mean squared percentage error (MAPE), and root mean squared error(RMSE), using a dataset from Power Grid Company of Bangladesh Ltd. (PGCB). Forecasting results show that MLP for both cases exhibits the best accuracy of 98.947% and 99.217% with the least RMSE, MAPE, and MAE error rates in comparison with other models. © 2023 IEEE.

关键词： Errors

来源：评论

学校读者我要写书评

暂无评论

GFlowNet Training by Policy Gradients 41

GFlowNet Training by Policy Gradients

引用

41st International Conference on Machine Learning, ICML 2024

作者： Niu, Puhua Wu, Shili Fan, Mingzhou Qian, Xiaoning Department of Electrical and Computer Engineering Texas A&M University College StationTX United States Department of Computer Science and Engineering Texas A&M University College StationTX United States Computational Science Intiative Brookhaven National Laboratory UptonNY United States

Generative Flow Networks (GFlowNets) have been shown effective to generate combinatorial objects with desired properties. We here propose a new GFlowNet training framework, with policy-dependent rewards, that bridges keeping flow balance of GFlowNets to optimizing the expected accumulated reward in traditional Reinforcement-Learning (RL). This enables the derivation of new policy-based GFlowNet training methods, in contrast to existing ones resembling value-based RL. It is known that the design of backward policies in GFlowNet training affects efficiency. We further develop a coupled training strategy that jointly solves GFlowNet forward policy training and backward policy design. Performance analysis is provided with a theoretical guarantee of our policy-based GFlowNet training. Experiments on both simulated and real-world datasets verify that our policy-based strategies provide advanced RL perspectives for robust gradient estimation to improve GFlowNet performance. Our code is available at: ***/niupuhua1234/GFN-PG. Copyright 2024 by the author(s)

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Coverage Hole Aware Data Routing for Mobile Sink Wireless Sensor Networks 14

Coverage Hole Aware Data Routing for Mobile Sink Wireless Se...

引用

14th IEEE International Conference on Electronics Information and Emergency Communication, ICEIEC 2024

作者： Wedaj, Fisseha Teju Hawbani, Ammar Wang, Xingfu Al-Gunid, Haithm M. Yilma, Getinet Farooq Qaisar, Muhammad Umar University of Science and Technology of China School of Computer Science and Technology Hefei China Shenyang Aerospace University School of Computer Science Shenyang China Adama Science and Technology University School of Electrical Engineering and Computing Adama Ethiopia Southern University of Science and Technology School of System Design and Intelligent Manufacturing Shenzhen China

ISBN: (纸本)9798350361889

In wireless sensor networks (WSNs), the coverage hole problem is one of the challenging problems that needs an effective solution. Data routing protocols in WSNs aim to disseminate the sensors' data to the central device with the minimum possible network resource usage. However, suppose there is a coverage hole in the middle of the source and destination nodes, and the source node doesn't have information about the hole. In that case, the data may choose the longer and more repetitive path to reach the destination, increasing resource usage, mainly energy and time. The condition is even more challenging when the network employs a mobile data collector. This paper proposes a Coverage Hole Aware Data Routing (CHADR) protocol for Mobile Sink WSNs that determines the shorter hole bypassing path in advance at the source node using the hole and data collector location information. CHADR evaluates the path efficiency using directional angle and expected hops count. After determining the path, data will be transmitted to the mobile sink using an Opportunistic routing technique. We proved CHADR performance with experimental simulations. © 2024 IEEE.

关键词： Routing protocols

来源：评论

学校读者我要写书评

暂无评论

The New Frontier in Media: AI-Driven Content Creation for Ad-Supported TV using Generative Adversarial Network 7

The New Frontier in Media: AI-Driven Content Creation for Ad...

引用

7th International Conference of computer and Informatics engineering, IC2IE 2024

作者： Ramagundam, Shashishekhar Karne, Niharika Comcast Software Development and Engineering in Content Discovery and Ai United States Troy University Masters in Computer Science AL. Department of Computer Science Montgomery Campus United States Dt Amtrak Sr. Principal Systems Engineer in C and Rs Platform Services United States University of South Alabama Masters in Electrical Engineering Department of Electrical Engineering United States

ISBN: (纸本)9798331505530

The use of Artificial intelligence (AI) technology in the content creation to produce creative aspects like editing, audience analysis, creating ideas, writing copy, etc. The major aim is to streamline and automate the process of content creation and convert it into a more efficient and effective way. There is a lack of transparency in the AI content production process and hence it does not exactly mimic human activities, which includes fantasizing and picking up new abilities. AI content creation requires more creativity while investigating with a specific goal in mind. Although, AI faces many challenges since it constantly expands its base knowledge. One of the biggest ethical problems is the probability that AI content is used to deceive or influence people. Hence, to overcome these difficulties a machine learning-based Artificial Intelligent content creation framework is generated for content creation in Ad-supported TV. Here, an effective model Generative Adversarial Network (GAN) introduced to deliver enterprise data, customize descriptions, and adjust content based on consumer behaviors. This developed model produces material that is closely related to the preferences of the user. AI content creation increases efficiency and productivity by saving time and money. It also leads to the development of better content. Finally, the experimental analysis is performed to find the effectiveness of the developed deep learning-based AI content creation framework via various metrics. © 2024 IEEE.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Heart Diseases Prediction Using Machine Learning Techniques: A Case Study from Jordan 25

Heart Diseases Prediction Using Machine Learning Techniques:...

引用

25th International Arab Conference on Information Technology, ACIT 2024

作者： Alazaidah, Raed Safaqah, Mohammed Abu Alzoubi, Haneen Omar, Hani Al-Safarini, Maram Y. Elsoud, Esraa Abu Elrashidi, Ali Alluwaici, Mo'Ath Owaida, Suhaila Abu Zarqa University Faculty of It Department of Data Science and Ai Zarqa Jordan Department of Mechanical Engineering Muscat Oman Applied Science Private University Faculty of It Department of Computer Science Amman Jordan Zarqa University Faculty of It Department of Computer Science Zarqa Jordan University of Business and Technology Department of Electrical Engineering Jeddah21432 Saudi Arabia Al-al-Bayt University Faculty of Information Technology Department of Computer Science Mafraq Jordan

ISBN: (纸本)9798331540012

Heart disease is one of the most common diseases in Jordan. It is a major reason of death among Jordanian adult citizens. Worldwide, an average of 56,000 people dies each day or one death every 1.5 seconds. Hence, this research is interested in the early prediction of this disease among Jordanian people. To achieve this main objective, Machine Learning (ML) is utilized through a large number of classification models and considering five well-known evaluation metrics. These classification models have been trained on a primary dataset that has been collected for the purpose of the research. The results revealed that RandomForest and RandomCommittee are the best classification models to handle the task of the early prediction of Heart disease in Jordan. © 2024 IEEE.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

A novel underwater image restoration method based on decomposition network and physical imaging model

引用

INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS 2022年第9期37卷 5672-5690页

作者： Cui, Yanfang Sun, Yujuan Jian, Muwei Zhang, Xiaofeng Yao, Tao Gao, Xin Li, Yiru Zhang, Yan School of Information and Electrical Engineering Ludong University Yantai China School of Information Science and Engineering Linyi University Linyi China School of Computer Science and Technology Shandong University of Finance and Economics Jinan China

Underwater image restoration is one of the significant research in marine engineering and aquatic robotics. However, due to the propagation characteristics of light and the serious turbidity in underwater, the captured images often have chromatic aberration and scattering blur, which brings great challenges to the restoration of the raw image. In this paper, a revised underwater imaging model is proposed first, which reanalyzes the generation of background light from the atmosphere to the underwater and provides important support for underwater color correction. And then a network framework via the revised model is designed, which can decompose the captured image into different components corresponding to the revised model. The proposed network consists of a decomposition architecture with residual blocks that learns a complete separation of clear image and transmittance features. These two features are used along with the raw image to predict the background light. Finally, combining three constraints of the imaging model, the proposed framework can converge rapidly along the desired direction. By comparison with the performance of the state-of-the-art algorithms, the designed network shows excellent visibility and is capable of removing water on both synthetic and real-world images in different water types.

关键词： decomposition architecture image restoration underwater imaging model

来源：评论

学校读者我要写书评

暂无评论

The comparison of EEG features in specific Mandarin words' speech

The comparison of EEG features in specific Mandarin words' s...

引用

2023 International Conference on Fuzzy Theory and Its Applications, iFUZZY 2023

作者： Tsai, Wen-Shan Huang, Syue-Guan Huang, Kuan-Chih Chen, Jung-Chih Lin, Chin-Teng National Yang Ming Chiao Tung University Institute of Biomedical Engineering College of Electrical and Computer Engineering Hsinchu300093 Taiwan National Yang Ming Chiao Tung University College of Electrical and Computer Engineering Department of Electronics and Electrical Engineering Hsinchu300093 Taiwan National Yang Ming Chiao Tung University Brain Science and Technology Center College of Electrical and Computer Engineering Hsinchu300093 Taiwan Faculty of Engineering and Information Technology UltimoNSW2007 Australia

ISBN: (纸本)9798350305791

Due to advancements in brain signal application technology, there has been a growing focus on the human speech Brain-computer Interface (BCI) in recent years. An essential initial phase in crafting speech recognition systems is comprehending the inherent features of nature speech processing. To ensure the decoding capability of human speech BCI, researchers have to investigate the differences between overt speech and imagined speech, and they also need to find the best paradigm to build a reliable BCI. Thus, this study compared the differences in brain dynamics between overt speech and imagined speech while using the four Mandarin words ""(up), ""(down), ""(left), and ""(right). We first calculated the EEG alpha and beta band power for multiple channels situated in the regions of Broca's and Wernicke's areas (specifically F5, F7, FC3, FC5, and C5), to analyze the brain activity changes while speaking or imagining speech. Second, we used Daubechies 4 (db4) wavelet to do a 7-level decomposition to the EEG signal. The corresponding discrete wavelet transform (DWT) coefficients were further calculated with six statistical features for support vector machine (SVM) classification. We found that the "speak (overt speech)"task and the "imagine (imagined speech)"task resulted in significantly different alpha band power changes, and the beta band shows the same phenomenon. The result of average classification accuracy is 71.97% for "speak"and "imagine"distinction. In conclusion, we analyzed the brain dynamics and find that overt speech and imagined speech could be differentiate through DWT coefficients features. This may pave the way to establish a Mandarin imagined speech BCI. © 2023 IEEE.

关键词： Electroencephalography

来源：评论

学校读者我要写书评

暂无评论

A Cost-Effective Autonomous Beach Cleaning Rover Using Deep Learning 1

引用

6th International Conference on Information Systems and Management science, ISMS 2023

作者： Girish, S. Harikrishnan, S. Nair, Rahul H. Sreerajan, Midhun Sathyan, Soumya Pandi, V. Ravikumar Valsan, Vipina Suresh, Kavya Department of Computer Science and Engineering Amrita School of Computing Amrita Vishwa Vidyapeetham Amritapuri India Department of Electrical and Electronics Engineering Amrita School of Engineering Amrita Vishwa Vidyapeetham Amritapuri India

ISBN: (数字)9783031664106

ISBN: (纸本)9783031664090

This work aims to transform beach garbage management by developing an autonomous rover that utilizes deep learning and computer vision. The main goal is to enable the rover to traverse coastal environments on its own while using its sophisticated onboard camera to identify various waste products. The rover will identify different sorts of rubbish by employing the YOLO algorithm that uses the Bounding Box Regression technique to improve waste management tactics. The work involves hardware design, neural network training, real-time computer vision integration, and extensive field testing. The hardware components of the rover are carefully engineered through iterative development, guaranteeing smooth mobility and efficient data collection. By incorporating real-time computer vision techniques, the rover will be able to evaluate visual data and make decisions that will help identify waste. In order to verify the project’s effectiveness in tackling beach pollution field testing is carried out to confirm the rover’s accuracy in navigation and waste detection. An autonomous rover that embodies technical innovation for long-term, environmentally sustainable preservation is successfully developed. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

关键词： Integration testing

来源：评论

学校读者我要写书评

暂无评论

Avian Activity Classification Using Recurrent Networks to Fuse Videos with Metadata on Imbalanced Datasets 24

Avian Activity Classification Using Recurrent Networks to Fu...

引用

10th International Conference on Computing and Artificial Intelligence, ICCAI 2024

作者： Wang, Xijun Szymanski, Adam Hamada, Yuki Katsaggelos, Aggelos K. Dept. of Computer Science Northwestern University EvanstonIL United States Argonne National Lab LemontIL United States Dept. of Electrical and Computer Engineering Northwestern University EvanstonIL United States

ISBN: (纸本)9798400717055

Activity classification plays a crucial role in various real-life scenarios involving both humans and animals. There is an increasing need for precise activity classification focused on avian-solar interactions, as the usage of solar energy facilities, such as photovoltaic array power stations, has been observed to impact bird species richness, behavior, and activity. However, there has been no work to develop an automated system to monitor and classify these avian-solar interactions. All current methods rely on human observers, which is time and human resources costly and subject to errors related to searcher efficiency. With the recent success of Deep Learning models in activity classification problems, this paper develops a recurrent neural network-based model to automatically classify six avian activities around solar energy facilities. Our proposed model integrates critical feature engineering metadata with video frame data, enabling improved learning and more accurate activity classification. Furthermore, we address the challenge of data imbalance during training and demonstrate the efficacy of our model in detecting and classifying different activities within video tracks. Additionally, we analyze the saliency/backpropagation map of the trained proposed model and validate its decision-making rationale. © 2024 Owner/Author.

关键词： Solar power generation

来源：评论

学校读者我要写书评

暂无评论

An Automated Deep Learning Based Muscular Dystrophy Detection and Classification Model

引用

computers, Materials & Continua 2022年第4期71卷 305-320页

作者： T.Gopalakrishnan Periakaruppan Sudhakaran K.C.Ramya K.Sathesh Kumar Fahd N.Al-Wesabi Manal Abdullah Alohali Anwer Mustafa Hilal School of Computer Science and Engineering Vellore Institute of TechnologyVellore632014India Department of Computer Science and Engineering SRM TRP Engineering CollegeTiruchirappalli621105India Department of Electrical and Electronics Engineering Sri Krishna College of Engineering and TechnologyCoimbatore641008India School of Computing Kalasalingam Academy of Research and EducationKrishnankoil626128India Department of Computer Science King Khalid UniversityMuhayel AseerKSA Faculty of Computer and IT Sana’a UniversityYemen Department of Information Systems College of Computer and Information SciencesPrincess Nourah Bint Abdulrahman UniversitySaudi Arabia Department of Computer and Self Development Preparatory Year DeanshipPrince Sattam bin Abdulaziz UniversityAlkharjSaudi Arabia

Muscular Dystrophy (MD) is a group of inherited muscular diseases that are commonly diagnosed with the help of techniques such asmuscle biopsy, clinical presentation, and Muscle Magnetic Resonance Imaging(MRI). Among these techniques, Muscle MRI recommends the diagnosis ofmuscular dystrophy through identification of the patterns that exist in musclefatty replacement. But the patterns overlap among various diseases whereasthere is a lack of knowledge prevalent with regards to disease-specific ***, artificial intelligence techniques can be used in the diagnosis ofmuscular dystrophies, which enables us to analyze, learn, and predict forthe future. In this scenario, the current research article presents an automated muscular dystrophy detection and classification model using SynergicDeep Learning (SDL) method with extreme Gradient Boosting (XGBoost),called SDL-XGBoost. SDL-XGBoost model has been proposed to act as anautomated deep learning (DL) model that examines the muscle MRI dataand diagnose muscular dystrophies. SDL-XGBoost model employs Kapur’sentropy based Region of Interest (RoI) for detection purposes. Besides, SDLbased feature extraction process is applied to derive a useful set of featurevectors. Finally, XGBoost model is employed as a classification approach todetermine proper class labels for muscle MRI data. The researcher conductedextensive set of simulations to showcase the superior performance of SDLXGBoost model. The obtained experimental values highlighted the supremacyof SDL-XGBoost model over other methods in terms of high accuracy being96.18% and 94.25% classification performance upon DMD and BMD respectively. Therefore, SDL-XGBoost model can help physicians in the diagnosis of muscular dystrophies by identifying the patterns of muscle fatty replacementin muscle MRI.

关键词： Muscle magnetic resonance imaging XGBoost synergic deep learning roI detection kapur’s entropy muscular dystrophies

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：