检索结果-内蒙古大学图书馆

12th International Conference on Learning Representations, ICLR 2024

作者： Ishikawa, Satoki Karakida, Ryo Department of Computer Science Tokyo Institute of Technology Japan Artificial Intelligence Research Center AIST Japan

Second-order optimization has been developed to accelerate the training of deep neural networks and it is being applied to increasingly larger-scale models. In this study, towards training on further larger scales, we identify a specific parameterization for second-order optimization that promotes feature learning in a stable manner even if the network width increases significantly. Inspired by a maximal update parameterization, we consider a one-step update of the gradient and reveal the appropriate scales of hyperparameters including random initialization, learning rates, and damping terms. Our approach covers two major second-order optimization algorithms, K-FAC and Shampoo, and we demonstrate that our parameterization achieves higher generalization performance in feature learning. In particular, it enables us to transfer the hyperparameters across models with different widths. © 2024 12th International Conference on Learning Representations, ICLR 2024. All rights reserved.

关键词： Parameterization

来源：评论

学校读者我要写书评

暂无评论

Diminishing Unclear Consequences of Missing Values in Data Mining 8th

Diminishing Unclear Consequences of Missing Values in Data M...

引用

8th International Conference on Information and Communication Technology for Competitive Strategies, ICTCS 2023

作者： Pareshbhai, Bhathawala Vaishnavi Buch, Sanjay H. Department of Computer Science Bhagwan Mahavir Center for Advance Research Bhagwan Mahavir University Surat India

ISBN: (纸本)9789819994854

In the realm of data mining, the presence of missing values poses significant challenges that can undermine the accuracy and reliability of analytical outcomes. This study delves into the critical task of addressing missing values to mitigate the potential for ambiguous results in data mining processes. Recognizing the pivotal role of complete and accurate data in generating meaningful insights, this article explores various approaches for handling missing values, including omission, imputation, interpolation, and model-based techniques with valuable insights into selecting the most appropriate strategy based on contextual factors. Study also provides information about the potential of model-based imputation with their variants. The research article highlights the nuanced process of model selection and its pros and cons. The study provides a layman framework that integrates both traditional and innovative methodologies;this study contributes to a holistic understanding of mitigating the impact of missing values. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： Approaches Data mining Imputation Missing values Omission

来源：评论

学校读者我要写书评

暂无评论

A Scalable Algorithm for Fair Influence Maximization with Unbiased Estimator

引用

IEEE Transactions on Knowledge and Data Engineering 2025年第7期37卷 3881-3895页

作者： Rui, Xiaobin Wang, Zhixiao Peng, Hao Chen, Wei Yu, Philip S. China University of Mining and Technology School of Computer Science Jiangsu Xuzhou221116 China Mine Digitization Engineering Research Center of the Ministry of Education Jiangsu Xuzhou221116 China Beihang University School of Cyber Science and Technology Beijing100191 China Theory Center of Microsoft Research Asia Beijing100080 China University of Illinois at Chicago Department of Computer Science ChicagoIL60607 United States

This paper studies the fair influence maximization problem with efficient algorithms. In particular, given a graph G, a community structure C consisting of disjoint communities, and a budget k, the problem asks to select a seed set S (||S|| = k) that maximizes the influence spread while narrowing the influence gap between different communities. This problem derives from some significant social scenarios, such as health interventions (e.g. suicide/HIV prevention) where individuals from racial minorities or LGBTQ communities may be disproportionately excluded from the benefits of the intervention. To depict the concept of fairness in the context of influence maximization, researchers have proposed various notions of fairness, where the welfare fairness notion that better balances fairness level and influence spread has shown promising effectiveness. However, the lack of efficient algorithms for optimizing the objective function under welfare fairness restricts its application to networks of only a few hundred nodes. In this paper, we modify the objective function of welfare fairness to maximize the exponentially weighted sum and the logarithmically weighted sum over all communities' influenced fractions (utility). To achieve efficient algorithms with theoretical guarantees, we first introduce two unbiased estimators: one for the fractional power of the arithmetic mean and the other for the logarithm of the arithmetic mean. Then, by adapting the Reverse Influence Sampling (RIS) approach, we convert the optimization problem to a weighted maximum coverage problem. We also analyze the number of reverse reachable sets needed to approximate the fair influence at a high probability. Finally, we present an efficient algorithm that guarantees 1 - 1/e - ϵ (positive objective function) or 1 + 1/e + ϵ (negative objective function) approximation for any small ϵ > 0. Experiments demonstrate that our proposed algorithm could efficiently handle large-scale networks with good performa

关键词： Approximation algorithms

来源：评论

学校读者我要写书评

暂无评论

Federated Learning for Enhanced Intrusion Detection in Smart City Environments 18

Federated Learning for Enhanced Intrusion Detection in Smart...

引用

18th International Conference on Open Source Systems and Technologies, ICOSST 2024

作者： Hamid, Soomaiya Bawany, Narmeen Zakaria Jinnah University for Women Center for Computing Research Department of Computer Science and Software Engineering Karachi Pakistan

ISBN: (纸本)9798331508692

The integration of IoT devices in smart cities enhances urban infrastructure, services, and governance but also introduces significant cybersecurity challenges. Traditional centralized Intrusion Detection Systems (IDS) face several issues, including data privacy concerns and high-power consumption due to centralized data processing. These challenges increase the risks of unauthorized access, data breaches, and privacy violations, undermining user trust and compliance with privacy regulations. Additionally, the centralization of data and processing leads to higher power consumption, making these systems less sustainable for widespread deployment in smart cities. This research addresses these issues by proposing a Federated Learning (FL)based intrusion detection framework for smart cities. FL enables collaborative and privacy-preserving model training across distributed IoT devices, mitigating the need to share sensitive data centrally. By aggregating local model updates, FL ensures data privacy and distributes the computational workload, significantly reducing power consumption compared to traditional centralized systems. The proposed model leverages advanced AI techniques and is trained using the IoTID20 dataset. The Flower framework, utilizing the FedAvg algorithm, facilitates the federated learning process. Our experimental results demonstrate that the global model achieves 98% accuracy, with individual clients achieving accuracies of around 85% to 98%. This approach provides continuous learning mechanisms, anomaly detection, and ensemble learning capabilities, enhancing the resilience of federated intrusion detection systems against emerging threats and adversarial attacks. This research systematically investigates the application of federated learning for intrusion detection in smart city networks, addressing key challenges and advancing the state-of-the-art in decentralized cybersecurity solutions. The proposed framework offers a robust, scalable, and privacyco

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Naïve Bayesian Approach to Intruder Detection in Smart Homes Using Sensor Networks 4th

Naïve Bayesian Approach to Intruder Detection in Smart Ho...

引用

4th International Conference on Advances in Computational science and Engineering, ICACSE 2023

作者： Bolivar, Daisy Jane S. Gom-os, Dhong Fhel K. Department of Computer Science Center for Research in Intelligent Systems University of the Philippines Cebu Cebu6000 Philippines

ISBN: (纸本)9789819729760

We used camera, light, door, and window sensors, and day of the week, duration, and activation time parameters for a Naïve Bayesian approach to intruder detection in a smart home. A nine-week normal and a two-week intruder behavior data were extracted using sensor devices. SmartThings SmartApp was used to gather the sensor data which were published to Google Sheets, then exported to Excel and Python for pre-processing, providing a dataset size of 509, consisting of 284 normal and 225 intruder data. The data had an 80/20 train-test split and was fitted to the Naïve Bayes model. By identifying behavioral patterns and increasing parameters from the state-of-the-art, the model’s accuracy improved to 94.12% and achieved an F2-score of 93.18%. This study has shown that increasing sensor size could significantly improve intruder detection. Future work could include the analysis of the contribution of each sensor to provide informed decision to system designers. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： Smart homes

来源：评论

学校读者我要写书评

暂无评论

Use that Pinky on the ‘A’ Key: A Finger-Key Identification Module for a Touch Typing Trainer 4th

Use that Pinky on the ‘A’ Key: A Finger-Key Identificati...

引用

4th International Conference on Advances in Computational science and Engineering, ICACSE 2023

作者： Valles, Oscar Vian L. Gom-os, Dhong Fhel K. Center for Research in Intelligent Systems Department of Computer Science University of the Philippines Cebu Cebu6000 Philippines

ISBN: (纸本)9789819729760

Typing tests list out words for the user to repeat. These tests measure metrics that quantify the user’s ability to type. However, conventional typing metrics do not measure correct finger placement. This aspect of typing may affect the user’s health. It is also a crucial aspect of keyboard typing education. As such, a method to identify which finger is used to press which key is beneficial. This paper introduces a new technique to achieve this by developing a finger-key identification module that utilizes computer vision algorithms to detect the keys in a keyboard, and a ready-made machine learning solution to track fingers while typing. This module was successful in finger-key identification with an accuracy of 99.58% in a dataset of 942 keypresses. It also had an average finger-key identification time of 0.083 s, which allows it to perform real-time finger-key identification for typing speeds up to 143.040 words per minute, well beyond the median typing speed of the population. Two metrics for correct finger placement were defined and were implemented in a proof-of-concept trainer: (1) finger placement accuracy and (2) historical finger placement accuracy. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： Palmprint recognition

来源：评论

学校读者我要写书评

暂无评论

Personality-based and trust-aware products recommendation in social networks

引用

Applied Intelligence 2023年第1期53卷 879-903页

作者： Vatani, Nasim Rahmani, Amir Masoud Javadi, Hamid Haj Seyyed Department of Computer Engineering Science and Research Branch Islamic Azad University Tehran Iran Future Technology Research Center National Yunlin University of Science and Technology 123 University Road Section 3 Yunlin Douliou64002 Taiwan Department of Mathematics and Computer Science Shahed University Tehran Iran

In recent years, with the development of technology, the shopping approach of people has moved towards pervasive online social shopping. As a result, how to create a recommendation algorithm that offers products based on the personal and different needs and tastes of people on social networks is a significant research issue. This article proposes a personality-based and trust-aware probabilistic product recommendation algorithm in social networks. We present a dynamic method for determining how similar people in social networks are. For this purpose, we consider the personality-based features of recommendation attributes of products in social networks. Then, the level of trust of products and types of correlations among the products is considered to create a probabilistic matrix of product recommendation. Moreover, for solving the cold start problem of products, we consider qualitative aspects of products while exploiting personality-based user behavior regarding their purchases. At last, the empirical experiments are conducted to analyze the impact of the algorithm’s different influence factors using the Amazon dataset. Moreover, the results of comprehensive experiments adopted to verify the proposed personalized recommendation algorithm’s effectiveness show that the proposed algorithm has the appropriate effectiveness and the higher accuracy. © 2022, The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature.

关键词： Behavioral research

来源：评论

学校读者我要写书评

暂无评论

SeerahGPT: Retrieval Augmented Generation based Large Language Model 18

SeerahGPT: Retrieval Augmented Generation based Large Langua...

引用

18th International Conference on Open Source Systems and Technologies, ICOSST 2024

作者： Obaid, Surayya Bawany, Narmeen Zakaria Center for Computing Research Department of Computer Science and Software Engineering Jinnah Univesity for Women Karachi Pakistan

ISBN: (纸本)9798331508692

Large Language Models(LLMs) have become widely recognized in recent years for their exceptional performance in language generation capabilities. As a result, an unprecedented rise is seen in its use cases in various domains specifically involving Natural Language Processing(NLP). These models however perform suboptimally when exploited in the field of studies where authenticity of the generated content is a critical aspect. One such domain is usage of LLM for exploration of the life of Prophet Muhammad S.A.W(commonly referred to as Seerah). It is of utmost significance to ensure the authenticity and reliability in the sources used and reported by the LLM due to the sensitive nature of the domain. The contemporary LLMs, however, lack the explainability in their response due to their inherent black-box nature. In our study, we have presented a novel LLM named SeerahGPT that addresses this challenge with the help of retrieval-augmented generation (RAG). This technique enables the model to utilize both parametric and nonparametric memories for generating response of queries. Our model, built on the Llama-2-7b architecture, employs Sentence Transformer embedding to effectively retrieve relevant information. The model's capabilities are augmented by integrating it with a corpus having Islamic texts such as the Quranic translation and Hadith collections, and historical accounts. The model's performance is benchmarked against its base model using both quantitative and qualitative metrics. The comparative analysis with Llama-2-7b revealed that SeerahGPT incorporation with external knowledge sources, provided more authentic and verifiable responses, despite the others exhibiting greater fluency. Performance metrics such as BLEU, ROUGE, and METEOR indicated SeerahGPT's better accuracy and contextual handling. This study paves way for analysis of such sensitive domains in more efficient way that can be utilized in other complex domains such as Islamic theology and Fiqh or legal

关键词： Natural language processing systems

来源：评论

学校读者我要写书评

暂无评论

Approximation of a Regulator for Discrete Two- Parameter Coupled Systems 17

Approximation of a Regulator for Discrete Two- Parameter Cou...

引用

17th International Conference on Management of Large-Scale System Development, MLSD 2024

作者： Danik, Yulia Dmitriev, Mikhail 'Computer Science and Control' of Russian Academy of Sciences Department 71 Federal Research Center Moscow Russia

ISBN: (纸本)9798350375718

In this paper, a stabilizing feedback law for a coupled discrete dynamic system with two parameters is proposed based on the use of asymptotic approximations in powers of two parameters and the decomposition of the original system. The proposed method uses the Pade approximations and the SDRE approach. Illustrative numerical experiments are presented. © 2024 IEEE.

关键词： Feedback

来源：评论

学校读者我要写书评

暂无评论

Dynamic Evolutionary Game-based Modeling,Analysis and Performance Enhancement of Blockchain Channels

引用

IEEE/CAA Journal of Automatica Sinica 2023年第1期10卷 188-202页

作者： PeiYun Zhang MengChu Zhou ChenXi Li Abdullah Abusorrah IEEE the School of Computer Science Nanjing University of Information Science&TechnologyNanjing 210044China the Helen and John C.Hartmann Department of Electrical and Computer Engineering New Jersey Institute of TechnologyNewarkNJ 07102 USA the School of Computer and Information Anhui Normal UniversityWuhu 241003China the Department of Electrical and Computer Engineering Faculty of Engineeringand Center of Research Excellence in Renewable Energy and Power SystemsKing Abdulaziz UniversityJeddah 21481Saudi Arabia

The recent development of channel technology has promised to reduce the transaction verification time in blockchain *** transactions are transmitted through the channels created by nodes,the nodes need to cooperate with each *** one party refuses to do so,the channel is unstable.A stable channel is thus *** nodes may show uncooperative behavior,they may have a negative impact on the stability of such *** order to address this issue,this work proposes a dynamic evolutionary game model based on node *** model considers various defense strategies'cost and attack success ratio under *** can dynamically adjust their strategies according to the behavior of attackers to achieve their effective *** equilibrium stability of the proposed model can be *** proposed model can be applied to general channel *** is compared with two state-of-the-art blockchain channels:Lightning network and Spirit *** experimental results show that the proposed model can be used to improve a channel's stability and keep it in a good cooperative stable *** its use enables a blockchain to enjoy higher transaction success ratio and lower transaction transmission delay than the use of its two peers.

关键词： Blockchain channel network evolutionary game malicious behavior secure computing stability analysis

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：