检索结果-内蒙古大学图书馆

1st IEEE Colombian Caribbean Conference, C3 2023

作者： Arrubla-Hoyos, Wilson Seveiche-Maury, Zurisaddai Saeed, Khalid Gomez, Jorge Eliecer Gomez De-la-Hoz-Franco, Emiro University of Sucre Faculty of Engineering Sincelejo Colombia Bialystok University of Technology Faculty of Computer Science Bialystok Poland University of Cordoba Faculty of Engineering Monteria Colombia Corporación Universidad de la Costa Department Computer Science and Electronics Barranquilla Colombia

ISBN: (纸本)9798350341799

Dengue disease, spread by mosquitoes, affects a large part of the world's population. Early diagnosis is essential to avoid its severe impacts. This paper seeks to compare classical machine learning techniques with ensemble approaches in the early classification of dengue: Dengue without alarm signs (DNWS), Dengue with alarm signs (DWWS) and Severe Dengue (SD). A dataset available at https://***/ from the Colombian government with 53,814 records and 38 attributes is used. The data are processed in Google Colab using Pandas, Matplotlib and Numpy, while Scikit-learn is used for modeling. The results are supported by a detailed confusion matrix, revealing the actual performance of each model and highlighting the superiority of the ensemble approaches over classical techniques in early dengue classification. The Stacking method achieves an accuracy (Acc) of 88%, outperforming all other techniques employed. Among the classical techniques, Decision Tree (DT) achieved the best result with an Acc of 84%. In conclusion, when contrasting the performance of classical techniques and ensemble approaches in early dengue classification, it is highlighted that the latter demonstrate robustness in terms of quality metrics, such as accuracy, recall and F1-Score. This observation suggests that ensemble approaches have the potential to overcome the limitations associated with classical techniques, resulting in more robust and accurate achievements. © 2023 IEEE.

关键词： Decision trees

来源：评论

学校读者我要写书评

暂无评论

FAIRSTREAM: FAIR MULTIMEDIA STREAMING BENCHMARK FOR REINFORCEMENT LEARNING AGENTS

arXiv

引用

arXiv 2024年

作者： Weil, Jannis Ringsdorf, Jonas Barthel, Julian Chen, Yi-Ping Phoebe Meuser, Tobias Technical University of Darmstadt Germany Department of Computer Science and Information Technology La Trobe University Australia

Multimedia streaming accounts for the majority of traffic in today’s internet. Mechanisms like adaptive bitrate streaming control the bitrate of a stream based on the estimated bandwidth, ideally resulting in smooth playback and a good Quality of Experience (QoE). However, selecting the optimal bitrate is challenging under volatile network conditions. This motivated researchers to train Reinforcement Learning (RL) agents for multimedia streaming. The considered training environments are often simplified, leading to promising results with limited applicability. Additionally, the QoE fairness across multiple streams is seldom considered by recent RL approaches. With this work, we propose a novel multi-agent environment that comprises multiple challenges of fair multimedia streaming: partial observability, multiple objectives, agent heterogeneity and asynchronicity. We provide and analyze baseline approaches across five different traffic classes to gain detailed insights into the behavior of the considered agents, and show that the commonly used Proximal Policy Optimization (PPO) algorithm is outperformed by a simple greedy heuristic. Future work includes the adaptation of multi-agent RL algorithms and further expansions of the environment. © 2024, CC BY.

关键词： Observability

来源：评论

学校读者我要写书评

暂无评论

MODELING THE HODGKIN-HUXLEY NEURON TO DETERMINE NEURONAL ENERGY CONSUMPTION EFFICIENCY AND OXYGEN CONSUMPTION VALUES

Biomedical Sciences Instrumentation

引用

Biomedical sciences Instrumentation 2022年第2期58卷 50-58页

作者： Enochson, Avery Ewert, Daniel Mitar, Dipankar Striker, Ryan Cleveland, Jerika Wolf, Henry Gieser, Anna Braaten, Ben Allen, Jeffery Allen, Monica Department of Integrated Engineering at Iron Range Engineering Program of Minnesota State University Mankato MN United States Department of Computer Science and Computer Engineering University of Wisconsin – La Crosse La CrosseWI United States Department of Electrical and Computer Engineering North Dakota State University FargoND United States Air Force Research Laboratory Munitions Directorate Eglin Air Force Base Eglin AFB FL United States

Understanding neuronal structure and function is essential to studying the human brain. The goal of this project was to create a model of human brain neurons that accurately reflects neuronal function, energy consumption, and oxygen consumption. Extensive work has been performed on the Hodgkin-Huxley model of neurons to accurately model neuronal firing. This study focuses on the creation of a model of the Hodgkin-Huxley neuron in MATlaB with the assistance of the DynaSim toolbox. This model was used to compute values and using another model the energy efficiency of the neuron was calculated and related to oxygen consumption, which then corresponds to Blood Oxygenation Level Dependent (BOLD) imaging in fMRI. The results of this study provide detailed visual and technical information about how the brain neurons function, which is crucial to the further development in brain imaging techniques and to further our understanding of why and how our complex brains function. © 2022 IAE All rights reserved.

关键词： Neurons

来源：评论

学校读者我要写书评

暂无评论

Causal Reasoning in large language Models using Causal Graph Retrieval Augmented Generation

Causal Reasoning in Large Language Models using Causal Graph...

引用

Conference on Human System Interactions, HSI

作者： Chamod Samarajeewa Daswin De Silva Evgeny Osipov Damminda Alahakoon Milos Manic Centre for Data Analytics and Cognition La Trobe University Australia Department of Computer Science Electrical and Space Engineering Luleå University of Technology Sweden Department of Computer Science Virginia Commonwealth University Richmond USA

ISBN: (数字)9798350362916

ISBN: (纸本)9798350362923

large language Models (LLMs) are leading the Generative Artificial Intelligence transformation in natural language understanding. Beyond language understanding, LLMs have demonstrated capabilities in reasoning tasks, including commonsense, logical, and mathematical reasoning. However, their proficiency in causal understanding has been limited due to the complex nature of causal reasoning. Several recent studies have discussed the role of external causal models for improved causal understanding. Building on the success of Retrieval-Augmented Generation (RAG) for factual reasoning in LLMs, this paper introduces a novel approach that utilizes Causal Graphs as external sources for establishing causal relationships between complex vectors. This method is empirically evaluated using two benchmark datasets across the metrics of Context Relevance, Answer Relevance, and Grounding, in its ability to retrieve relevant context with causal alignment. The retrieval effectiveness is further compared with traditional RAG methods that are based on semantic proximity.

关键词： Measurement Limiting large language models Semantics Benchmark testing Cognition Vectors

来源：评论

学校读者我要写书评

暂无评论

A Comparative Study of Optimized-LSTM Models Using Tree-Structured Parzen Estimator for Traffic Flow Forecasting in Intelligent Transportation

引用

computers, Materials & Continua 2025年第5期83卷 3369-3388页

作者： Hamza Murad Khan Anwar Khan Santos Gracia Villar Luis Alonso DzulLopez Abdulaziz Almaleh Abdullah M.Al-Qahtani Department of Electronics University of PeshawarPeshawar25120KPKPakistan Department of Project Management(L.A.D.L) Higher Polytechnic School(S.G.V)Universidad Europea del AtlanticoSantander39011Spain Faculty of Engineering Universidad Internacional IberoamericanaCampeche24560Mexico Faculty of Engineering Universidade Internacional do CuanzaCuitoEN250BieAngola Department of Project Management Universidad Internacional IberoamericanaCampeche24560Mexico Department of Project Management Department of Management SciencesUniversid de La RomanaLa Romana22000Dominican Reublic Information Systems Department College of Computer ScienceKing Khalid UniversityAbha61421Saudi Arabia Department of Electrical and Electronic Engineering College of Engineering&Computer ScienceJazan UniversityJazan45142Saudi Arabia

Traffic forecasting with high precision aids Intelligent Transport Systems(ITS)in formulating and optimizing traffic management *** algorithms used for tuning the hyperparameters of the deep learning models often have accurate results at the expense of high computational *** address this problem,this paper uses the Tree-structured Parzen Estimator(TPE)to tune the hyperparameters of the Long Short-term Memory(LSTM)deep learning *** Tree-structured Parzen Estimator(TPE)uses a probabilistic approach with an adaptive searching mechanism by classifying the objective function values into good and bad *** ensures fast convergence in tuning the hyperparameter values in the deep learning model for performing prediction while still maintaining a certain degree of *** also overcomes the problem of converging to local optima and avoids timeconsuming random search and,therefore,avoids high computational complexity in prediction *** proposed scheme first performs data smoothing and normalization on the input data,which is then fed to the input of the TPE for tuning the *** traffic data is then input to the LSTM model with tuned parameters to perform the traffic *** three optimizers:Adaptive Moment Estimation(Adam),Root Mean Square Propagation(RMSProp),and Stochastic Gradient Descend with Momentum(SGDM)are also evaluated for accuracy prediction and the best optimizer is then chosen for final traffic prediction in TPE-LSTM *** results verify the effectiveness of the proposed model in terms of accuracy of prediction over the benchmark schemes.

关键词： Short-term traffic prediction sequential time series prediction TPE tree-structured parzen estimator LSTM hyperparameter tuning hybrid prediction model

来源：评论

学校读者我要写书评

暂无评论

Managing health insurance using blockchain technology

arXiv

引用

arXiv 2023年

作者： Ananna, Tajkia Nuri Saifuzzaman, Munshi Chowdhury, Mohammad Jabed Morshed Ferdous, Md Sadek Department of Computer Science and Engineering Metropolitan University Sylhet Bangladesh Dynamic Solution Innovators Ltd Dhaka Bangladesh Department of Computer Science and IT La Trobe University Australia Department of Computer Science and Engineering BRAC University Bangladesh

Health insurance plays a significant role in ensuring quality healthcare. In response to the escalating costs of the medical industry, the demand for health insurance is soaring. Additionally, those with health insurance are more likely to receive preventative care than those without health insurance. However, from granting health insurance to delivering services to insured individuals, the health insurance industry faces numerous obstacles. Fraudulent actions, false claims, a lack of transparency and data privacy, reliance on human effort and dishonesty from consumers, healthcare professionals, or even the insurer party itself, are the most common and important hurdles towards success. Given these constraints, this chapter briefly covers the most immediate concerns in the health insurance industry and provides insight into how blockchain technology integration can contribute to resolving these issues. This chapter finishes by highlighting existing limitations as well as potential future directions. Copyright © 2023, The Authors. All rights reserved.

关键词： Blockchain

来源：评论

学校读者我要写书评

暂无评论

FedSC: Provable Federated Self-supervised Learning with Spectral Contrastive Objective over Non-i.i.d. Data 41

FedSC: Provable Federated Self-supervised Learning with Spec...

引用

41st International Conference on Machine Learning, ICML 2024

作者： Jing, Shusen Yu, Anlan Zhang, Shuai Zhang, Songyang Department of Radiation Oncology University of California San FranciscoCA United States Department of Electrical and Computer Engineering Lehigh University BethlehemPA United States Department of Data Science New Jersey Institute of Technology NewarkNJ United States Department of Electrical and Computer Engineering University of Louisiana at Lafayette LafayetteLA United States

Recent efforts have been made to integrate self-supervised learning (SSL) with the framework of federated learning (FL). One unique challenge of federated self-supervised learning (FedSSL) is that the global objective of FedSSL usually does not equal the weighted sum of local SSL objectives. Consequently, conventional approaches, such as federated averaging (FedAvg), fail to precisely minimize the FedSSL global objective, often resulting in suboptimal performance, especially when data is non-i.i.d.. To fill this gap, we propose a provable FedSSL algorithm, named FedSC, based on the spectral contrastive objective. In FedSC, clients share correlation matrices of data representations in addition to model weights periodically, which enables inter-client contrast of data samples in addition to intra-client contrast and contraction, resulting in improved quality of data representations. Differential privacy (DP) protection is deployed to control the additional privacy leakage on local datasets when correlation matrices are shared. We also provide theoretical analysis on the convergence and extra privacy leakage. The experimental results validate the effectiveness of our proposed algorithm. Copyright 2024 by the author(s)

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Short Wave Infrared Upconversion Imager with Electron Blocking layer to Screen Electron Injection into Photosensitive layer 5

Short Wave Infrared Upconversion Imager with Electron Blocki...

引用

5th IEEE International Conference on Flexible and Printable Sensors and Systems, FLEPS 2023

作者： Shin, Chanho Li, Ning Ng, Tse Nga University of California San Diego Material Science and Engineering Program La Jolla United States University of California San Diego Department of Electrical and Computer Engineering La Jolla United States

ISBN: (纸本)9781665457330

In this work, we demonstrated upconversion imagers integrated with shortwave infrared photodetectors paired with an electron blocking layer. The use of electron blocking layer screened charge injection to prevent recombination in photosensitive layer. The characteristics of each electron blocking layer were analyzed in aspects of noise and detectivity. For the optimized device, the parasitic luminance in the dark was efficiently suppressed, and the photon-to-photon efficiency was increased. The electron blocking layer used in this work is generally applicable for upconversion imagers using different absorption and emitting materials. © 2023 IEEE.

关键词： Photodetectors

来源：评论

学校读者我要写书评

暂无评论

Filling the SME credit gap: a systematic review of blockchain-based SME finance literature

Journal of Trade Science

引用

Journal of Trade science 2023年第2/3期11卷 45-72页

作者： Deepak Kumar B.V. Phani Naveen Chilamkurti Suman Saurabh Vanessa Ratten Department of Industrial and Management Engineering Indian Institute of Technology Kanpur Kanpur India Department of Computer Science and Information Technology La Trobe University Melbourne Australia Department of Management and Marketing La Trobe University Melbourne Australia

Purpose The review examines the existing literature on blockchain-based small and medium enterprise (SME) finance and highlights its trend, themes, opportunities and challenges. Based on these factors, the authors create a framework for the existing literature on blockchain-based SME financing and lay down future research paths. Design/methodology/approach The review follows a systematic approach. It includes 53 articles encompassing multiple dimensions of blockchain-based SME finance, including peer-to-peer lending platforms, supply chain finance (SCF), decentralized lending protocols and tokenization of assets. The review critically evaluates these approaches' theoretical underpinnings, empirical evidence and practical implementations. Findings The review demonstrates that blockchain-based SME finance holds significant promise in addressing the credit gap by leveraging blockchain technology's decentralized and transparent nature. Benefits identified include reduced information asymmetry, improved access to financing, enhanced credit assessment processes and increased financial inclusion. However, the literature acknowledges several challenges and limitations, such as regulatory uncertainties, scalability issues, operational complexities and potential security risks. Originality/value The article contributes to the growing knowledge of blockchain-based SME finance by synthesizing and evaluating the existing literature. It also provides a framework for the existing literature in the area and future research paths. The study offers insights for researchers, policymakers and practitioners seeking to understand the potential of blockchain technology in filling the SME credit gap and fostering economic development through improved access to finance for SMEs.

关键词： SME MSME Credit gap Blockchain technology Supply chain finance Blockchain technology adoption DeFi ICOs STOs

来源：评论

学校读者我要写书评

暂无评论

Efficient k-means with Individual Fairness via Exponential Tilting

arXiv

引用

arXiv 2024年

作者： Zhu, Shengkun Zeng, Jinshan Sun, Yuan Wang, Sheng Li, Xiaodong Peng, Zhiyong The School of Computer Science Wuhan University China The School of Computer and Information Engineering Jiangxi Normal University China La Trobe Business School La Trobe University Australia The School of Computing Technologies RMIT University Australia

In location-based resource allocation scenarios, the distances between each individual and the facility are desired to be approximately equal, thereby ensuring fairness. Individually fair clustering is often employed to achieve the principle of "treating all points equally", which can be applied in these scenarios. This paper proposes a novel algorithm, tilted k-means (TKM), aiming to achieve individual fairness in clustering to ensure fair allocation of resources. We integrate the exponential tilting into the sum of squared errors (SSE) to formulate a novel objective function called tilted SSE. We demonstrate that the tilted SSE can generalize to SSE and employ the coordinate descent and first-order gradient method for optimization. We propose a novel fairness metric, the variance of the squared distance of each point to its nearest centroid within a cluster, which can alleviate the Matthew Effect typically caused by existing fairness metrics. Our theoretical analysis demonstrates that the well-known k-means++ incurs a multiplicative error of O(k log k) with our objective function, and we establish the convergence of TKM under mild conditions. In terms of fairness, we prove that the variance in each cluster generated by TKM decreases with t, where t is a hyperparameter that adjusts the trade-off between utility and fairness. In terms of efficiency, we demonstrate the time complexity of TKM is linear with the dataset size. Moreover, we demonstrate the monotonicity of the tilted SSE with respect to t in a simple case. Our experimental results demonstrate that TKM outperforms state-of-the-art methods in effectiveness, fairness, and efficiency. Specifically, TKM exhibits a better trade-off between clustering utility and fairness than six baselines and achieves hundreds or even thousands of times acceleration in running time. Moreover, TKM can overcome the RAM overflow issue that other methods encounter with a large dataset size. © 2024, CC BY.

关键词： Resource allocation

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：