检索结果-内蒙古大学图书馆

International Conference on Artificial Intelligence in Information and Communication (ICAIIC)

作者： Anh-Tien Tran Thanh Phung Truong Dang-Huy Mac Dongwook Won Jaemin Kim Nhu-Ngoc Dao Sungrae Cho School of Computer Science and Engineering Chung-Ang University Seoul Korea South Department of Computer Science and Engineering Sejong University Seoul Korea South

ISBN: (数字)9798331507022

ISBN: (纸本)9798331507039

Maximizing energy efficiency (EE) in Multiple-Input Single-Output (MISO) downlink networks employing Quadrature Rate-Splitting Multiple Access (Q-RSMA) is a challenging task due to the non-convex optimization problem involving power allocations, beamforming vectors, and rate allocations under multiple constraints. In this paper, we propose a Deep Reinforcement Learning (DRL) framework based on the Deep Deterministic Policy Gradient (DDPG) algorithm to maximize EE. We handle the minimum rate constraints by formulating the rate allocation as a linear programming (LP) problem, which allows for a computationally efficient solution. The beamforming vector normalization is clarified to ensure the unit norm constraints are satisfied. Simulation results demonstrate the effectiveness of the proposed approach in achieving high EE while satisfying all system constraints.

关键词： Array signal processing Simulation Downlink Deep reinforcement learning Linear programming Energy efficiency Vectors Resource management MISO Optimization

来源：评论

学校读者我要写书评

暂无评论

FairEval: Evaluating Fairness in LLM-Based Recommendations with Personality Awareness

arXiv

引用

arXiv 2025年

作者： Sah, Chandan Kumar Lian, Xiaoli Xu, Tony Zhang, Li School of Computer Science and Engineering Beihang University Beijing China Department of Electrical and Computer Engineering McGill University Montreal Canada

Recent advances in Large Language Models (LLMs) have enabled their application to recommender systems (RecLLMs), yet concerns remain regarding fairness across demographic and psychological user dimensions. We introduce FairEval, a novel evaluation framework to systematically assess fairness in LLM-based recommendations. Unlike prior benchmarks that focus solely on demographic attributes, FairEval uniquely integrates personality traits with eight sensitive demographic attributes, including gender, race, and age enabling a comprehensive and nuanced assessment of user-level bias. We evaluate state-of-the-art models, including ChatGPT 4o and Gemini 1.5 Flash, on music and movie recommendation tasks using structured prompts. FairEval’s personality-aware fairness metric, PAFS@25, achieves high consistency scores up to 0.9969 for ChatGPT 4o and 0.9997 for Gemini 1.5 Flash, underscoring its robustness in equitable recommendations across diverse user profiles, while also uncovering fairness gaps, with SNSR disparities reaching up to 34.79%. Our results also reveal disparities in recommendation consistency across user identities and prompt formulations, including typographical and multilingual variations. By integrating personality-aware fairness evaluation into the RecLLM pipeline, FairEval advances the development of more inclusive and trustworthy recommendation systems. © 2025, CC BY-SA.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Prediction of Domestic Wastewater Discharge in Urban Towns of Pearl River Based on ARIMA 9th

Prediction of Domestic Wastewater Discharge in Urban Towns o...

引用

9th International Conference on Advances in Energy Resources and Environment engineering, ICAESEE 2023

作者： Liu, Lidong He, Yantao Liu, Sijia Liu, Suping Luo, Yongbin Mo, Jiayu Liu, Chunzhen Department of Computer Science Guangdong University of Science and Technology Dongguan523668 China School of Mechanical Engineering Hunan University of Science and Technology Xiangtan411201 China School of Teacher Education Shanwei Institute of Technology Shanwei516600 China

ISBN: (纸本)9783031665042

The discharge of domestic sewage by urban residents has always been a major factor causing water pollution in rivers, and with the continuous improvement of people's living standards, the amount of domestic sewage discharge is also increasing year by year. In order to predict the specific situation of domestic sewage discharge by urban residents in the Pearl River area in the future, this paper establishes an ARIMA model, combined with the existing 21 year domestic sewage discharge data, to make predictions for the next 10 years. During the prediction process, the original data is subjected to ADF test to obtain the optimal difference data, and this data is subjected to ACF and PACF analysis, and the model test table is drawn. The white noise sequence of the data was examined using both ACF and PACF. The analysis concluded that both AR residuals and MA residuals are sequences of white noise, thereby meeting the prerequisites for the ARIMA model. Subsequently, the formula for the ARIMA model was derived, and a model diagram of the time prediction sequence was illustrated. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： River pollution

来源：评论

学校读者我要写书评

暂无评论

The Missing Data Imputation Method for Medical Data Classification Using Most Probable Value

The Missing Data Imputation Method for Medical Data Classifi...

引用

International Conference on Communication Systems and Network Technologies (CSNT)

作者： Balasaheb Tarle Vaishali Tidake Rupali Tajanpure Ratnesh Litoriya Department of Computer Engineering MVP Samaj's KBTCOE Nashik India Department of Computer Science and Engineering Medi-Caps University Indore India

ISBN: (数字)9798331531935

ISBN: (纸本)9798331531942

In medical data classification, improving the performance of classification is essential when the dataset's size is small, and some attributes' values need to be included. Classification performance can suffer if the training dataset starts with insufficient data samples. The challenges like overfitting, high variance, inability to find real performance of model, biased models, sensitivity to noise, improper evaluation occurs with small dataset. The proposed algorithm uses domain-based multiple imputations to create additional data tuples and appends these tuples to the originally available training data. This improves the classification power of the classifier. This proposed technique uses a set of domain values to impute missing attribute values. The accuracy of imputed tuples is verified using classifier filters. The proposed technique significantly improves the classification performance of the classifier. This technique is suitable for small to medium-sized data sets. Data imputation helps develop improved and more accurate classifiers. The proposed approach offers improved classification performance. The proposed method achieves a 4.01% improvement in the classification accuracy of fuzzy neural network FBNFC. We performed two hypothesis tests, the t-test, and the Wilcoxon rank sum test. It is observed that the alternative hypothesis is true.

关键词： Training Accuracy Filters Sensitivity Communication systems Noise Training data Fuzzy neural networks Imputation Overfitting

来源：评论

学校读者我要写书评

暂无评论

The Effect of Class Distribution on Multi-Modal Medical Images Classification in Meta-Learning

The Effect of Class Distribution on Multi-Modal Medical Im...

引用

Joint IAPR International Workshops on Statistical Techniques in Pattern Recognition and Structural and Syntactic Pattern Recognition, S+SSPR 2024

作者： Almugbel, Zainab Computing Department Applied College Imam Abdulrahman Bin Faisal University Dammam Saudi Arabia Computer Science and Engineering School of Innovation Design and Engineering Mälardalen University Västerås Sweden

ISBN: (纸本)9783031805066

Medical images are available in small datasets with different modalities and for various organs. Although Transfer learning is a promising approach for training models on small datasets, further studies are required on using them with several image modalities and body parts. This paper explores two dominant meta-learning algorithms: a metric-based algorithm, namely Prototypical Network, and an optimization-based algorithm, namely MAML. The algorithms trained on a small multi-modal medical dataset (i.e., Slake), and a general dataset of the same size (i.e., Tiny-imagenet) with different class distribution methods: Random, Logical-based, and Statistical-based class distribution. The aim is to apply diversity among the classes in the meta-training set and similarity between the classes in the meta-training, and meta-testing or meta-validation sets. The results validate the importance of class distribution on the accuracy of the algorithms. MAML is hard to be trained on Tiny-imagenet but it shows good accuracy on Slake in specific cases. In statistical distribution, the Prototypical Network shows high accuracy on the datasets, especially when the similarity between the meta-training and meta-validation sets is considered for Tiny-imagenet. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Analysis of Challenges of IoT Adoption for Smart City Application 5th

Analysis of Challenges of IoT Adoption for Smart City Applic...

引用

5th International Conference on Innovative Product Design and Intelligent Manufacturing Systems, ICIPDIMS 2023

作者： Das, Rashmi Prava Muduli, Debendra Luhach, Ashish Kumar Department of Computer Science and Engineering C V Raman Global University Bhubaneswar752054 India School of Electrical and Communication Engineering Papua New Guinea University of Technology Lae411 Papua New Guinea

ISBN: (纸本)9789819767311

The Internet of Things (IoT) is an architecture for a network that processes and analyzes sensitive data in order to deliver a variety of services to a large number of users. IoT devices are capable of delivering a wide variety of value-added services in a variety of smart city applications. Nevertheless, there have been a lot of worries voiced concerning data privacy, confidentiality, and security in relation to the IoT, as well as consumer resistance to adopting such innovative technology. This article examines difficulties linked to the expansion of the IoTs and their adoption by users in smart city applications using Step-wise Weight Assessment Ratio Analysis (SWARA) technique. A total of 30 factors that inhibit IoT adoption were analyzed and categorized under 5 major factors. It revealed that, functional and technological barriers offer more resistance to IoT adoption for smart city application in India. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： IoT Optimization Smart city SWARA

来源：评论

学校读者我要写书评

暂无评论

Unsupervised Domain Adaptation for Occlusion Resilient Human Pose Estimation

arXiv

引用

arXiv 2025年

作者： Dutta, Arindam Bose, Sarosij Bachu, Saketh Ta, Calvin-Khang Karydis, Konstantinos Roy-Chowdhury, Amit K. The Department of Electrical and Computer Engineering University of California Riverside United States The Department of Computer Science and Engineering University of California Riverside United States

Occlusions are a significant challenge to human pose estimation algorithms, often resulting in inaccurate and anatomically implausible poses. Although current occlusion-robust human pose estimation algorithms exhibit impressive performance on existing datasets, their success is largely attributed to supervised training and the availability of additional information, such as multiple views or temporal continuity. Furthermore, these algorithms typically suffer from performance degradation under distribution shifts. While existing domain adaptive human pose estimation algorithms address this bottleneck, they tend to perform suboptimally when the target domain images are occluded, a common occurrence in real-life scenarios. To address these challenges, we propose OR-POSE : Unsupervised Domain Adaptation for Occlusion Resilient Human POSE Estimation. OR-POSE is an innovative unsupervised domain adaptation algorithm which effectively mitigates domain shifts and overcomes occlusion challenges by employing the mean teacher framework for iterative pseudo-label refinement. Additionally, OR-POSE reinforces realistic pose prediction by leveraging a learned human pose prior which incorporates the anatomical constraints of humans in the adaptation process. Lastly, OR-POSE avoids overfitting to inaccurate pseudo labels generated from heavily occluded images by employing a novel visibility-based curriculum learning approach. This enables the model to gradually transition from training samples with relatively less occlusion to more challenging, heavily occluded samples. Extensive experiments show that OR-POSE outperforms existing analogous state-of-the-art algorithms by ∼ 7% on challenging occluded human pose estimation datasets. © 2025, CC BY.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Development of an Advanced Multi-Disease Prediction Framework Utilizing Machine Learning Techniques

Development of an Advanced Multi-Disease Prediction Framewor...

引用

Disruptive Technologies (ICDT), International Conference on

作者： Sudhanshu Maurya Arnav Kotiyal Atulya Gupta Nilesh Shelke Abhishek Bhattacherjee Tiyas Sarkar Symbiosis Institute of Technology Nagpur Campus Symbiosis International (Deemed University) School of Computer Science and Engineering Pune India Department of Computer Science and Engineering Graphic Era Deemed to be University Dehradun India Department of Computer Science and Engineering SRM Institute of Science and Technology Ghaziabad India School of Computer Science and Engineering Lovely Professional University Phagwara India

ISBN: (数字)9798331519582

ISBN: (纸本)9798331519599

This innovative study applies Machine Learning (ML) techniques to construct an all-encompassing Multiple Disease Prediction System using datasets from Kaggle. The dataset holds great potential for marking “Parkinson”, “Heart Disease” and “Heart diseases” causing to its multifaceted structure. The study incorporates Support Vector Machines for “Support Vector Machine” that organized Mikey's and problem of Diabetes predictions & Logistic regression for heart disease predictions. This novel approach means that the system offers improved reliability and ease of use, and most importantly, optimal predictions for every possible medical nuance. A prominent concern is the intersection of emerging technologies and medicine with emphasis on the importance of evidence-based medicine for primary healthcare. The project uses intuitive and advanced algorithms for ML based on the machine learning algorithms to improve the value of understanding diabetes for cardiovascular disease prediction and mitigate the gap for improved healthcare action. The principal expects to develop models which are economical and can support a wide range of changes, such as using decision trees, but utilizing more advanced techniques like SVM and Logistic Regression. The integration of the open source Streamlit platform allows for the user module of the integrated component to be more easily utilized. This improves operational efficacy and access to the predictive system, thus broadening the target user population. This integrates the different aspects of the project in an enhanced way to build a better and more proactive health care system for early diagnosis and preventive measures of common health problems.

关键词： Support vector machines Heart Logistic regression Machine learning algorithms Biological system modeling Medical services Machine learning Predictive models Diabetes Medical diagnostic imaging

来源：评论

学校读者我要写书评

暂无评论

A Multimodal Fusion and Ensemble Approach for Robust Fake News Detection through Deep Learning, Reinforcement Learning, and Using Blockchain to Enhance Security and Authorization

Recent Advances in Computer Science and Communications

引用

Recent Advances in computer science and Communications 2025年第2期18卷

作者： Kumar, Vivek Satveer Ahmad, Waseem Kumar, Satendra Department of Computer Application Quantum University U.K Roorkee India Department of Computer Science & Engineering Meerut Insitute of Engineering and Technology U.P Meerut India

Background: computer science significantly influences modern culture, especially with the rapid breakthroughs and technology in social media networking. Social media platforms have become significant channels for sharing and exchanging daily news and information on many issues in the current digital environment, which is known for its massive data collection and transmission capabilities. While there are many benefits to this environment, there are also many false reports and information that deceive readers and users into believing they are receiving correct information. Objective: Nowadays, all users use social media to obtain news content, but sometimes some malicious users tamper with real news and then spread fake news, which may reduce the reputation of social media. Therefore, many existing models have been introduced to detect fake news, but these models are based on traditional machine learning algorithms, such as decision tree (DT), multilayer perceptron (MLP), random forest (RF), etc. These models Lack of performance, security, and authorization. Our proposed model can solve existing model problems using reinforcement learning and blockchain technology. Methods: In this research paper, we explain a new way to identify fake news. This new approach, combined with policy-based heuristic reinforcement learning (PHRL), where the model dynamically adjusts through iterative learning, is the key innovation and gradually improves classification accuracy. The same as our smart contract authorization method, which enhances the authenticity of content posted safely by authorized users and improves the transparency and accountability of information. Results: Our model was tested on real-time information collected from various sources with 70% accuracy and valid authentication. Conclusion: Our proposed model produced better results with a Mean Absolute Error (MAE) of 0.0811 and Root Mean Squared Error (RMSE) of 0.2847, both significantly lower values. Our proposed mode

关键词： Deep reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Two-Stage Vision Transformer-Based Framework for Anomaly Detection and Classification in Surveillance Videos

Two-Stage Vision Transformer-Based Framework for Anomaly Det...

引用

International Conference on Electrical, computer and Communication engineering (ECCE)

作者： Mahedi Hasan Jubair Ahmed Nabin Naeem Mia Fahim Shakil Tamim Suzad Mohammad Dipta Mohon Das Department of Computer Science and Engineering IUBAT—International University of Business Agriculture and Technology Dhaka Bangladesh Department of Computer Science and Engineering Chittagong University of Engineering and Technology (CUET) Chattogram Bangladesh

ISBN: (数字)9798350357509

ISBN: (纸本)9798350357516

Anomaly detection and classification play a vital role in maintaining public safety and security. An automated system for anomaly detection and classification can reduce human leverage, cost, and time. We propose a two-stage classifier pipelined within a single model for anomaly detection and classification. The first stage of the two-stage classifier is a Convolutional Neural Network(CNN) based binary classifier that determines whether an event is anomalous or normal. Based on the output of the first classifier, if the event is found anomalous, then it goes to the second stage of our single-pipelined model. The second stage classifier is a Vision Transformer (ViT) based architecture that further classifies an anomalous event into specific anomaly categories. This research utilized the UCF Crime dataset. Being quite large, it requires a significant amount of computational resources and time for processing. We also proposed a keyframe extraction algorithm to reduce the computational cost and time. The proposed keyframe extraction algorithm identifies and selects only relevant frames from videos and discards the redundant and irrelevant frames. The proposed methodology combines Convolution Neural Network (CNN) and Vision Transformer (ViT) for spatial-temporal feature extraction from a complex scenario and classify them. The proposed model achieves 98% accuracy for binary classification modules and 95% accuracy for multi-class classification. Furthermore, the proposed keyframe extraction algorithm significantly reduces the processing time and computational resources. For each videos, it requires only 20ms processing time. The outcome of the proposed model suggests that it can outperform traditional methods for anomaly detection and classification. However, a highly correlated and vast amount of data creates problems like overfitting and increases the complexity of the model.

关键词： computer vision Accuracy Computational modeling Surveillance Transformers Feature extraction Classification algorithms Convolutional neural networks Anomaly detection Videos

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：