检索结果-内蒙古大学图书馆

International CSI computer Conference, CSICC

作者： Amir Sanati Mansoureh A. Dashtestani Habib Rostami Saeed Talatian Azad Computer Engineering Department Faculty of Intelligent System Engineering and Data Science Persian Gulf University Bushehr Iran

Pneumothorax is a life-threatening and urgent chest disease than can be detected using Chest X-Ray (CXR) image. CXR images are low resolution and diagnosis of pneumothorax based on them is error prone. Deep learning-based computer aided diagnosis systems can improve diagnosis performance of pneumothorax. Convolutional Neural Networks (CNNs) are default networks in deep learning-based medical image process. However, CNNs fail to capture long range features. On the other side, Transformer are proposed to exploit long range feature, but they cannot capture local features. In this paper, we propose a general method with a convolution and a transformer module which can classify CXR images to diagnose pneumothorax by extracting local features, global features and global features attended by local ones using a novel architecture. Results show that the proposed method outperforms base architectures and the other previous works.

关键词： Image resolution Convolution Neural networks computer architecture Feature extraction Transformers Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Study on Distributed Multi-Level Security Scheme For Cloud Computing

Study on Distributed Multi-Level Security Scheme For Cloud C...

引用

International Conference on Inventive Computation Technologies (ICICT)

作者： G Sathishkumar. P Siva. D Anto Maria Joseph. R. Santhosh R. Reshma Department of Computer Science and Engineering Faculty of Engineering Karpagam Academy of Higher Education Coimbatore Tamil Nadu India

Cloud computing seems to be a scalable and realistic distributed computing platform that delivers resources as a service. One of the major difficulties impeding cloud computing development is data security in the cloud. The security vulnerabilities in cloud computing and other cloud services slow down cloud computing growth and difficulties with data protection and privacy. In this new context, cloud service customers must be cautious about data breaches. This article studies a solution to current diverse security challenges that have arisen as a result of the service demonstrating that the offered methods are provably safe and very efficient. The solution comprises a distributed multi-level security scheme for processing the information involved in cloud computing. The proposed scheme evaluates the most significant encryption techniques generally used in various industries like AES (Advanced Encryption Standard), Blowfish, 3DES (Data Encryption Standard), CAST128, and RC4 (Rivest Cipher 4). The proposed model with these encryption techniques ensure minimal energy consumption in cloud computing with maximum speed. Moreover, to increase the security of the data, various cryptographical algorithms are applied, analyzed, and compared in an OpenStack environment.

关键词：

来源：评论

学校读者我要写书评

暂无评论

FireSense: AI-Based Wildfire and Smoke Detection in Forests

FireSense: AI-Based Wildfire and Smoke Detection in Forests

引用

International Conference on New Trends in Computing sciences (ICTCS)

作者： Abdul Basit Mughal Syed Mujtaba Haider Rafi Ullah Khan Syed Mohsin Ali Shah Javiriya Hameed Arain Department of Computer Science Bahria University Islamabad Pakistan Department of Mathmatics University of Pavia Pavia Italy Dept. of Electrical & Computer Engineering PAF-IAST Haripur Pakistan Faculty of Engineering University of Deusto Bilbao Spain Department of Computer Science COMSATS University Islamabad Pakistan

ISBN: (数字)9798331523411

ISBN: (纸本)9798331523428

Forests are an essential part of the environment which plays a vital role in sustaining ecological balance and providing habitat to different wildlife. Wildfires are one of the serious threats to forests which create a devastating impact on the natural ecosystem. Early identification of smoke and wildfires will help to reduce damage and support a quick reaction. Earlier, wildfire and smoke detection in forests were very difficult due to the challenging forest terrain, low light conditions, and difficult weather conditions. With the development in technology, several computer vision methods are combined with satellite and Unmanned Aerial Vehicle (UAV) data to correctly identify wildfires, but accuracy is still a concern with these models. This paper introduced a novel method to detect wildfire and smoke in forests utilizing fine-tuned YOLOv11n using an aerial vehicle dataset. The mAP of the fine-tuned YOLOv11n is 88.6%, with a precision of 100% and an F1-score of 85%. The number of parameters has decreased by 30% compared to its previous models, and the model size is only 5.5MB.

关键词： Wildfires Forests Satellites Biological system modeling Computational modeling Wildlife Habitats Ecosystems Autonomous aerial vehicles Data models

来源：评论

学校读者我要写书评

暂无评论

Classification under strategic self-selection 24

Classification under strategic self-selection

引用

Proceedings of the 41st International Conference on Machine Learning

作者： Guy Horowitz Yonatan Sommer Moran Koren Nir Rosenfeld Faculty of Computer Science Technion — Israel Institute of Technology Department of Industrial Engineering and Management Ben Gurion University

When users stand to gain from certain predictive outcomes, they are prone to act strategically to obtain predictions that are favorable. Most current works consider strategic behavior that manifests as users modifying their features; instead, we study a novel setting in which users decide whether to even participate (or not), this in response to the learned classifier. Considering learning approaches of increasing strategic awareness, we investigate the effects of user self-selection on learning, and the implications of learning on the composition of the self-selected population. Building on this, we propose a differentiable framework for learning under self-selective behavior, which can be optimized effectively. We conclude with experiments on real data and simulated behavior that complement our analysis and demonstrate the utility of our approach.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Novel Hybrid AI Framework for Real-Time Prediction and Risk Assessment of Vector-Borne Epidemics using Spatiotemporal Data

A Novel Hybrid AI Framework for Real-Time Prediction and Ris...

引用

Mobile Networks and Wireless Communications (ICMNWC), IEEE International Conference on

作者： Rajesh Munirathnam Meenakshi Maindola Bhargav H K Prachi Ramesh Patil Pooja Bhatt Natrayan L Data & Analytics Division TechnoGen Inc. New Jersey USA Department of Computer Science & Engineering Graphic Era Deemed to be University Dehradun India Department of Computer Science and Engineering Shridevi Institute of Engineering and Technology Industry Institute Interaction Cell(IIIC) Tumakuru India Department of Electronics and Telecommunication Engineering Jagdambha College of Engineering and Technology Yavatmal India Department of Computer science and Engineering Faculty of Engineering and Technology Parul institute of Engineering and Technology Parul University Vadodara India Department of Mechanical Engineering Saveetha School of Engineering SIMATS Chennai India

ISBN: (数字)9798331528348

ISBN: (纸本)9798331528355

Vector-borne epidemics constitute major threats to global health that call for improved modeling systems for managing diseases. In order to incorporate such factors into an actual model, a traditional forecasting process may fail to grasp the nature of disease transmission adequately. We present a hybrid AI system based on LSTM and GCN for forest bug outbreak detection in real time. The framework also integrated a spatial-temporal analysis of climate variability, vector population fluctuations, and human mobility with climate variables across 15 regions over a five-year period (2019–2023). The hybrid framework also produced superior results, achieving a 94.2% prediction rate for health risk disease outbreaks, compared to a 17.3% improvement in traditional machine learning. When used to distinguish between high-risk and low-risk zones, the model demonstrated the following metrics: sensitivity of 92.8% and specificity of 95.1%. Real-time risk assessment demonstrated 89.6% effectiveness in indicating early warning and the best prediction lead time average of 21 days. This framework significantly enhances the prediction of vector-borne diseases by offering a dependable, real-time platform for disease surveillance and risk assessment. Its high accuracy and early warning capabilities make the system a useful tool in epidemiological and public health campaigns and control measures.

关键词： Epidemics Accuracy Surveillance Real-time systems Vectors Spatiotemporal phenomena Risk management Public healthcare Long short term memory Diseases

来源：评论

学校读者我要写书评

暂无评论

Business Blockchain Suitability Determinants: Decision-Making through an Intuitionistic Fuzzy Method

引用

computer Systems science & engineering 2023年第11期47卷 1665-1690页

作者： Tomader Almeshal Jawad Berri Tarifa Almulhim Areej Alhogail Emam Ahmed Department of Information Systems College of Computer and Information SciencesKing Saud UniversityRiyadh11451Saudi Arabia Department of Quantitative Methods School of BusinessKing Faisal UniversityAl-Ahsa31982Saudi Arabia Faculty of Computer Science and Engineering King Salman International UniversitySouth SinaiEgypt

Blockchain is one of the innovative and disruptive technologies that has a wide range of applications in multiple industries beyond *** widespread adoption of blockchain technology in various industries has shown its potential to solve challenging business problems,as well as the possibility to create new business models which can increase a firm’s *** to the novelty of the technology,whereby many companies are still exploring potential use cases,and considering the complexity of blockchain technology,which may require huge changes to a company’s existing systems and processes,it is important for companies to carefully evaluate suitable use cases and determine if blockchain technology is the best solution for their specific *** research aims to provide an evaluation framework that determines the important dimensions of blockchain suitability assessment by identifying the key determinants of suitable use cases in a business *** this paper,a novel approach that utilizes both qualitative(Delphi method)and quantitative(fuzzy set theory)methods has been proposed to objectively account for the uncertainty associated with data collection and the vagueness of subjective *** work started by scanning available literature to identify major suitability dimensions and collected a range of criteria,indicators,and factors that had been previously identified for related *** opinions were then gathered using a questionnaire to rank the importance and relevance of these elements to suitability ***,the data were analyzed and we proceeded to integrate multi-criteria group decision-making(MCGDM)and intuitionistic fuzzy set(IFS)*** findings demonstrated a high level of agreement among experts,with the model being extremely sensitive to variances in expert ***,the results helped to refine and select the most relevant suitability determinants under three important dimensions:fu

关键词： Blockchain business blockchain use cases suitability assessment multi-criteria group decision-making fuzzy Delphi intuitionistic fuzzy set

来源：评论

学校读者我要写书评

暂无评论

Child Tracking and Prediction of Violence on Children In Social Media Using Natural Language Processing and Machine Learning 1

引用

22nd International Conference on Artificial Intelligence and Soft Computing, ICAISC 2023

作者： Nallakaruppan, M.K. Srivastava, Gautam Gadekallu, Thippa Reddy Reddy, Praveen Kumar Krishnan, Sivarama Polap, Dawid School of Information Technology and Engineering Vellore Institute of Technology Vellore India Department of Mathematics and Computer Science Brandon University Brandon Canada School of Information Technology and Engineering Vellore Institute of Technology Department of Electrical and Computer Engineering Lebanese American University Lebanon Zhongda Group Zhejiang Jiaxing314312 China Faculty of Applied Mathematics Silesian University of Technology Kaszubska 23 Gliwice44-100 Poland

ISBN: (数字)9783031425059

ISBN: (纸本)9783031425042

Crimes against children are a direct threat to the world of tomorrow. The future of the world is in the hands of children who are the real wealth of mankind. The challenge for society is to address and handle challenges against children such as sexual abuse, trafficking, child labour, and harassment. Even though measures are taken by several governments globally, violence against children is an evident day-to-day process. This becomes the most challenging issue in the post-Internet era, where organized crime against children takes place around the world through a targeted group of people through virtual private networks and the dark web. What comes next is the emergence of social media, which provides an opportunity for these criminals to harm children globally through powerful social networks [15]. The existing methods focus on classifiers that do not work across multiple keywords [6] to look for sentiment analysis regarding a single classifier. There are only a few works that address the issue and these methods do not employ the multi-classification model [2]. In this paper, we propose a model which includes the Twitter API, the APIFY web scraper framework, as well as the VADER semantic analyzer in Python which possesses Decision tree analysis of sentiment prediction. The proposed work provides 99.6904% accuracy and predicts the classification of sentiment analysis. Our work has novelty in attribute selection, child trade, and violence analysis in social media and prediction of classification accuracy using the Decision Tree algorithm. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

Detecting Hallucinations in Large Language Models Using Machine Learning Classifiers: A Comparative Study

Detecting Hallucinations in Large Language Models Using Mach...

引用

Computational Intelligence Approaches and Applications (ICCIAA), International Conference on

作者： Hazem Al-Najjar Nadia Al-Rousan Douaa Al-Rousan Department of Computer Science College of Information Technology University of Petra Amman Jordan Department of computer engineering College of Engineering AIBalqa Applied University Salt Jordan Department of Computer Engineering Faculty of Electrical Engineering and Information Systems German Jordanian University Amman Jordan School of Civil Engineering Universiti Sains Malaysia Nibong Tebal Pulau Pinang Malaysia

ISBN: (数字)9798331523657

ISBN: (纸本)9798331523664

This study is concerned with the comprehensive evaluation of machine learning models intended for the detection of hallucination in Large Language Model output. The preprocessing of the prompt and answers using TF-IDF vectorization has been done as a preliminary step before developing a hallucination prediction model. Afterwards, the generated dataset was then divided into training and testing, with percentages equal to 70% and 30%, respectively. The selected machine learning models were trained with the dataset created. In this research, eight various classifiers are implemented and compared. To evaluate the results of our approach, multiple metrics were taken into consideration: accuracy, precision, recall, and Fl-score. While the training results showed exceptional performance with accuracies exceeding 99% in several models, the test results indicated huge challenges in generalization. Random Forest had the highest testing accuracy of 94.77% at 65% precision but with a very low recall of 4.85%. Besides, cross validation is used with random forest and decision tree, the results showed that the cross-validation is better than the test and train splitting model. The study shows the significant trade-offs between precision and recall in hallucination detection, which is especially problematic while working with imbalanced datasets.

关键词： Training Measurement Accuracy Large language models Computational modeling Predictive models Decision trees Random forests Computational intelligence Testing

来源：评论

学校读者我要写书评

暂无评论

Heuristic-based Reasoning on Financial Knowledge Graphs

Heuristic-based Reasoning on Financial Knowledge Graphs

引用

2022 Workshops of the EDBT/ICDT Joint Conference, EDBT/ICDT-WS 2022

作者： Baldazzi, Teodoro Benedetto, Davide Brandetti, Matteo Vlad, Adriano Bellomarini, Luigi Sallinger, Emanuel Università Roma Tre Department of Computer Science and Engineering Rome Italy TU Wien Faculty of Informatics Vienna Austria University of Oxford Department of Computer Science Oxford United Kingdom Banca d’Italia Italy

When reasoning over large knowledge graphs, Datalog+/- languages offer a good trade-off between expressive power and computational complexity. However, in case of considerably large inputs and in the presence of recursive rules, even state-of-the-art reasoners struggle to accomplish reasoning tasks, e.g. answering Boolean Conjuctive Queries (BCQs). To address this problem, we introduce the notion of heuristic-based reasoning, that is, evaluating Datalog+/- rules according to an order of the facts determined via a user-defined heuristic. We show that adopting a fitting heuristic that guides the reasoning process provides an optimized way to answer a BCQ. To achieve this behaviour in practice, we enrich Datalog+/- programs with a context-aware operator, which we name Dynamic Hint Operator. We apply our new methodology to efficiently solve the Close Link problem on the knowledge graph of Italian companies, a relevant financial problem that estimates the risk to grant a specific loan to a company that is backed by collateral issued by another company. © 2022 Copyright for this paper by its authors.

关键词： Knowledge graph

来源：评论

学校读者我要写书评

暂无评论

Assessing the Effectiveness of Machine Learning Algorithms in Vehicle Insurance Fraud Prediction: A Comparative Approach

Assessing the Effectiveness of Machine Learning Algorithms i...

引用

Electronics and Renewable Systems (ICEARS), International Conference on

作者： Aditya P. Satone Subodh Daronde Department of Computer Science and Medical Engineering Faculty of Engineering and Technology Datta Meghe Institute of Higher Education and Research (DU) Wardha Maharashtra India Department of Biomedical Engineering Faculty of Engineering and Technology Datta Meghe Institute of Higher Education and Research (DU) Wardha Maharashtra India

ISBN: (数字)9798331509675

ISBN: (纸本)9798331509682

Due to the wealth of information accessible on many insurance issues, including travel and auto insurance, individuals are becoming more and more impacted by these kinds of insurance in today's society. But this tendency also makes room for dishonest practices. Insurance fraud is harmful and unlawful, regardless of who commits it—the vendor or the customer. Fraud can come from both the customer and the vendor. The consumer may commit fraud by filing exaggerated claims or using out-of-date policies, while the vendor may market insurance from phoney firms or neglect to submit premiums. This paper presents a comparative analysis of various classification algorithms, including Support Vector Machine (SVM), Random Forest (RF), Decision Tree (DT), Linear Regression (LR), and Polynomial Regression, for detecting insurance fraud. The performance of these algorithms is evaluated using key metrics such as Precision, Recall, and F1-Score. The analysis reveals that the bestperforming model is the Random Forest, with a Mean Squared Error (MSE) of 6.50, outperforming the other techniques.

关键词： Support vector machines Analytical models Machine learning algorithms Insurance Predictive models Data science Data models Fraud Classification algorithms Random forests

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：