检索结果-内蒙古大学图书馆

Frontiers Technology of Information and computer (ICFTIC), IEEE International Conference on

作者： Nan Wang Yang Liu School of Computer Science and Software Engineering University of Science and Technology Liaoning Anshan China

The field of sequential recommendation plays a crucial role in personalized recommendation systems, aiming to model users’ past interactions and predict their future interactions with items or behaviors. Traditional methods in sequential recommendation typically rely on user behavior history and item attributes for making recommendations, but they overlook the internal relationships and contextual information among items within a sequence. Moreover, existing autoencoder models face limitations in capturing long-term dependencies and effectively modeling contextual information for sequential recommendation tasks. To address these issues, we propose a novel framework called miSAASRec that leverages a multi-information autoencoder with a self-attention module to capture internal relevance and contextual features of the data. This enables miSAASRec to achieve more accurate and comprehensive data encoding, thereby enhancing the performance of the autoencoder. We conducted several experiments to demonstrate the superior performance of our miSAASRec model compared to existing methods, achieving improvements ranging from 7.92% and 32.20% in MRR (Mean Reciprocal Rank) and 9.00% and 28.90% in Recall@10.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Detecting Malicious Websites by using Deep Q-Networks

Detecting Malicious Websites by using Deep Q-Networks

引用

Cybersecurity Conference (SVCC), Silicon Valley

作者： Khanh Nguyen Younghee Park Software Engineering College of Engineering San Jose State University San Jose USA Computer Engineering College of Engineering San Jose State University San Jose CA

ISBN: (数字)9798350383140

ISBN: (纸本)9798350383157

Web applications are the most popular and critical applications for the Internet, but attackers also utilize web browsers for malicious purposes. Uniform Resource Loca- tor(URL) has been utilized to launch attacks on users, systems, and networks in different ways, such as phishing, defacement, malware, and spam. Identifying malicious URLs is the fundamental task to protect our digital assets. This paper presents a new malicious link detection system based on reinforcement learning to identify malicious URLs. It utilizes a Deep Q-Network (DQN) algorithm to make intelligent determinations by an agent in the designated environment. This paper analyzes malicious URL datasets with various feature sets with the DQN algorithm and provides a strategy to optimize the detection rates. The experimental results demonstrate the feasibility of preventing malicious URLs in different settings.

关键词： Uniform resource locators Training Adaptation models Accuracy Phishing Training data Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Diving into Continual Ultra-fine-grained Visual Categorization

Diving into Continual Ultra-fine-grained Visual Categorizati...

引用

2023 International Conference on Digital Image Computing: Techniques and Applications, DICTA 2023

作者： Zhang, Pengcheng Yu, Xiaohan Bai, Xiao Zheng, Jin Wu, Xiaoyu Gao, Yongsheng Beihang University School of Computer Science and Engineering State Key Laboratory of Software Development Environment Jiangxi Research Institute Beijing China Griffith University Institute for Integrated and Intelligent Systems Nathan4111 Australia

ISBN: (纸本)9798350382204

Recent advance in ultra-fine-grained visual categorization (ultra-FGVC) has significantly boosted the capability of deep neural networks for ultra-FGVC tasks. However, building models for continually learning to recognize increasing ultra-fine-grained categories is still under-explored. This limits the application of ultra-FGVC techniques in real-world production. To this, we take the first attempt for continual ultra-FGVC. By evaluating existing continual learning methods on the constructed continual ultra-FGVC benchmark, we observe that the main bottleneck lies in the limited model plasticity for incrementally adapting to new tasks. This can be caused by excessive anti-forgetting constraints as the difficult ultra-FGVC task requires substantial update of parameters, and over-fitting on early tasks given that the ultra-fine-grained categories are with very few training samples. To tackle these problems, we propose a joint self-supervised learning and prompting model. The prompt-based continual learning framework offers proper anti-forgetting operation by fixed pretrained vision transformer and adaptive prompt selection. By jointly optimizing the learnable prompts with an adversarial self-supervised loss, the over-fitting on each continual learning task is mitigated. Extensive experiments demonstrate that the proposed method outperforms existing continual learning methods on the challenging continual ultra-FGVC problem. © 2023 IEEE.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Stress Analysis in Social Media in Bengali language using BERT-based approaches

Stress Analysis in Social Media in Bengali language using BE...

引用

Computing, Applications and Systems (COMPAS), IEEE International Conference on

作者： Afroza Khanom Nargis Akter Sabrina Jahan Maisha Abdul Kader M. Masum Department of Computer Science & Engineering International Islamic University Chittagong Chattogram Bangladesh Department of Software Engineering Daffodil International University Dhaka Bangladesh

ISBN: (数字)9798331529765

ISBN: (纸本)9798331529772

Recent researches highlight a rapid increase in mental health issues, signaling a concerning rise in stress-related conditions such as high blood pressure, psoriasis, polycystic ovary syndrome (PCOS), etc., and these should have a sophisticated remedy that plays a big role in stress analysis. To proceed with that the ongoing connection between increased stress levels and social media has gained a lot of attention. This study focused on inspecting a complex language, Bengali by extracting comments from YouTube to analyze stress to provide a meaningful solution. We are aware of no research on stress analysis in the Bengali language, although many studies have been done on the same topic in other languages. Thus by paying closer attention to stress analysis in the Bengali language, our work suggested a BERT-based method that has shown good results on a range of NLP tasks. Our approach included Bangla-BERT-base and hybrid models that combine deep and transfer learning, like BERT-LSTM, BERT-GRU, and BERT-CNN-BiLSTM. Our task performed well with an accuracy of 90.36% in stress analysis utilizing the Bengali dataset. Our experiment also helped to get an understanding that hybrid models had an influence in increasing the performance of the BERT-base model. Though our research has made little contribution to the BNLP field considering its vast aspect, expanding the Bengali dataset and conducting further multi-class stress analysis work can have a remarkable contribution.

关键词： Analytical models Video on demand Social networking (online) Computational modeling Psoriasis Transfer learning Human factors Mental health Blood pressure Web sites

来源：评论

学校读者我要写书评

暂无评论

Monte Carlo forecasting of time series data using Polynomial-Fourier series model

引用

International Journal of Modeling, Simulation, and Scientific Computing 2021年第3期12卷 112-124页

作者： Salim Jibrin Danbatta Asaf Varol Software Engineering Department Firat University Elazig CenterElazigTurkey Computer Engineering Department Maltepe University College of Engineering and Natural Sciences Maltepe University34857 Maltepe/IstanbulTurkey

The perishable nature of tourism products and services makes forecasting an important tool for tourism planning,especially in the current COVID-19 pandemic *** forecast assists tourism organizations in decision-making regarding resource allocations to avoid *** study is motivated by the need to model periodic time series with linear and nonlinear trends.A hybrid Polynomial-Fourier series model that uses the combination of polynomial and Fourier fittings to capture and forecast time series data was *** proposed model is applied to monthly foreign visitors to Turkey from January 2014 to August 2020 dataset and diagnostic checks show that the proposed model produces a statistically good *** improve the model forecast,a Monte Carlo simulation scheme with 100 simulation paths is applied to the model *** mean of the 100 simulation paths within±2σbounds from the model curve was taken and found to give statistically acceptable results.

关键词： Time series forecasting Monte Carlo Fourier series polynomial COVID-19.

来源：评论

学校读者我要写书评

暂无评论

Hybrid Parallel Inference for Large Model on Heterogeneous Clusters for High Throughput 29

Hybrid Parallel Inference for Large Model on Heterogeneous C...

引用

29th IEEE International Conference on Parallel and Distributed Systems, ICPADS 2023

作者： Xu, Chaonong Kong, Weiming Liu, Min Zhang, Mingming Li, Chao Gong, Luqi Beijing Key Lab of Petroleum Data Mining Beijing China Institute of Intelligent Computing Zhejiang Lab Hangzhou China Bupt School of Computer Science National Pilot Software Engineering School Beijing China

ISBN: (纸本)9798350330717

In high-throughput intelligent computing scenarios, multi-device parallelism strategies based on data parallelism or pipeline parallelism have been extensively utilized to accelerate large deep neural network model inference. Data parallelism offers nearly linear improvement in inference speed, but it is limited by the memory capacity of a single device which constrains the model size. On the other hand, pipeline parallelism can support larger models, but the total communication of the activations among devices is high, which limits the improvement of the inference speed. To address the demand for efficient model inference in high-throughput heterogeneous scenarios, we proposes a hybrid parallelism strategy that combines data parallelism and pipeline parallelism. The strategy involves grouping heterogeneous device clusters and then employing inter-group data parallelism along with intra-group pipeline parallelism. Moreover, we propose an algorithm to find an optimal hybrid parallel inference strategy with maximum throughput. The control variables of the strategy includes the number of groups, group-device assignments and model partition ratios. Our experimental evaluation demonstrates that compared to PipeEdge, a pipeline parallel inference framework for heterogeneous cluster, our strategy can achieve 1.7× -3.4× acceleration in an 8-device heterogeneous cluster without loss of accuracy. © 2023 IEEE.

关键词： data parallelism heterogeneous cluster hybrid parallelism model inference pipeline parallelism

来源：评论

学校读者我要写书评

暂无评论

Deep Word Embedding and Language Model based Sentimental Analysis of COVID-19 Tweets

Deep Word Embedding and Language Model based Sentimental Ana...

引用

International Bhurban Conference on Applied Sciences & Technology, IBCAST

作者： Sobia Qaiser M. Usman Akram Dept. of Computer and Software Engineering National University of Sciences and Technology Islamabad Pakistan

ISBN: (数字)9798350308259

ISBN: (纸本)9798350308266

The great Covid-19 pandemic affected billions of people's lives personally and socially. The research involves the analysis of public's views and opinions shared on Twitter social media platform related to Covid-19 pandemic and its detrimental or non-detrimental effects on public's mental health by using machine learning algorithm. The main purpose of this research includes analyzing public views related to Covid-19 pandemic by classifying the Tweets collected from the Twitter social platform. The proposed approach combines deep word embedding with MiniLM as an encoder to produce word vectors of high-dimensionality to preserve the words' semantic information. The resultant word vectors were used to train the model for the classification of the tweet in five sentiments i.e. Positive, Extremely Positive, Negative, Extremely Negative, and Neutral. The methodology is tested using publicly available Kaggle dataset as well as privately collected tweets. The comparative evaluation of the models revealed that MiniLM outperformed existing BERT based counterparts and attained highest accuracy of 93% with the Kaggle dataset. This analysis can assist the medical health authorities to monitor health information, conduct, and plan interventions to lower the pandemic effect and can help government to take precautionary measures.

关键词： COVID-19 Machine learning algorithms Pandemics Social networking (online) Blogs Semantics Government Mental health Vectors Monitoring

来源：评论

学校读者我要写书评

暂无评论

Explainable Artificial Intelligence (XAI) from a user perspective: A synthesis of prior literature and problematizing avenues for future research

引用

Technological Forecasting and Social Change 2023年第PartA期186卷

作者： Haque, AKM Bahalul Islam, A.K.M. Najmul Mikalef, Patrick Department of Software Engineering LUT University Finland Department of Computer Science Norwegian University of Science and Technology Norway

The rapid growth and use of artificial intelligence (AI)-based systems have raised concerns regarding explainability. Recent studies have discussed the emerging demand for explainable AI (XAI);however, a systematic review of explainable artificial intelligence from an end user's perspective can provide a comprehensive understanding of the current situation and help close the research gap. The purpose of this study was to perform a systematic literature review of explainable AI from the end user's perspective and to synthesize the findings. To be precise, the objectives were to 1) identify the dimensions of end users' explanation needs;2) investigate the effect of explanation on end user's perceptions, and 3) identify the research gaps and propose future research agendas for XAI, particularly from end users' perspectives based on current knowledge. The final search query for the Systematic Literature Review (SLR) was conducted on July 2022. Initially, we extracted 1707 journal and conference articles from the Scopus and Web of Science databases. Inclusion and exclusion criteria were then applied, and 58 articles were selected for the SLR. The findings show four dimensions that shape the AI explanation, which are format (explanation representation format), completeness (explanation should contain all required information, including the supplementary information), accuracy (information regarding the accuracy of the explanation), and currency (explanation should contain recent information). Moreover, along with the automatic representation of the explanation, the users can request additional information if needed. We have also described five dimensions of XAI effects: trust, transparency, understandability, usability, and fairness. We investigated current knowledge from selected articles to problematize future research agendas as research questions along with possible research paths. Consequently, a comprehensive framework of XAI and its possible effects on user behavio

关键词： Transparency

来源：评论

学校读者我要写书评

暂无评论

Dissect Black Box: Interpreting for Rule-Based Explanations in Unsupervised Anomaly Detection 38

Dissect Black Box: Interpreting for Rule-Based Explanations ...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Zhang, Yu Li, Ruoyu Wu, Nengwu Li, Qing Lin, Xinhan Hu, Yang Li, Tao Jiang, Yong Shanghai Artificial Intelligence Laboratory China Peng Cheng Laboratory China College of Computer Science and Software Engineering Shenzhen University China Tsinghua University China Tsinghua Shenzhen International Graduate School China Hunan University of Science and Technology China

In high-stakes sectors such as network security, IoT security, accurately distinguishing between normal and anomalous data is critical due to the significant implications for operational success and safety in decision-making. The complexity is exacerbated by the presence of unlabeled data and the opaque nature of black-box anomaly detection models, which obscure the rationale behind their predictions. In this paper, we present a novel method to interpret the decision-making processes of these models, which are essential for detecting malicious activities without labeled attack data. We put forward the Segmentation Clustering Decision Tree (SCD-Tree), designed to dissect and understand the structure of normal data distributions. The SCD-Tree integrates predictions from the anomaly detection model into its splitting criteria, enhancing the clustering process with the model's insights into anomalies. To further refine these segments, the Gaussian Boundary Delineation (GBD) algorithm is employed to define boundaries within each segmented distribution, effectively delineating normal from anomalous data points. At this point, this approach addresses the curse of dimensionality by segmenting high-dimensional data and ensures resilience to data variability and perturbations through flexible boundary fitting. We transform the intricate operations of anomaly detection into an interpretable rule's format, constructing a comprehensive set of rules for understanding. Our method's evaluation on diverse datasets and models demonstrates superior explanation accuracy, fidelity, and robustness over existing method, proving its efficacy in environments where interpretability is paramount. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Method for Improving the Performance of Linear Interpolation Implementation using the Formation of Two-step Movements

Method for Improving the Performance of Linear Interpolation...

引用

International Conference on Advanced computer Information Technologies (ACIT)

作者： Olexandr Romanyuk Yevhen Zavalniuk Oksana Romanyuk Anatoliy Snigur Oleksandr Reshetnik Volodymyr Maidaniuk Department of Software Engineering Vinnytsia National Technical University Vinnytsia Ukraine Department of Computer Engineering Vinnytsia National Technical University Vinnytsia Ukraine

The purpose of the work is increasing the productivity of evaluation function-based line segment formation through the usage of two-step movements and parallelization of movements calculation.

ISBN: (数字)9798350350043

ISBN: (纸本)9798350350050

The purpose of the work is increasing the productivity of evaluation function-based line segment formation through the usage of two-step movements and parallelization of movements calculation.

关键词： Surveys Productivity computers Interpolation Differential equations software Computational efficiency Information technology

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：