检索结果-内蒙古大学图书馆

32nd Irish Conference on Artificial Intelligence and Cognitive science, AICS 2024

作者： Pakrashi, Arjun Rushe, Ellen Namee, Brian Mac Brennan, Rob Yaman, Bezya School of Computer Science University College Dublin BelfieldDublin 4 Ireland School of Computer Science Dublin City University Collins Ave Ext Whitehall Dublin 9 Ireland The Insight Centre for Data Analytics Ireland SFI Centre for Research Training in Machine Learning Ireland ADAPT Centre Ireland School of Computer Science Trinity College Dublin College Green Dublin 2 Ireland

The 32nd Irish Conference on Artificial Intelligence and Cognitive science (AICS 2024), hosted by University College Dublin (UCD) in collaboration with Dublin City University (DCU), featured high-quality research in AI and cognitive science, with a focus on trustworthy AI, fairness, and bias mitigation. Of 82 submissions, 44 papers were accepted, covering applications in healthcare, smart mobility, financial forecasting, anomaly detection, and sports analytics, alongside fundamental machine learning advancements. Keynote speakers Dr. Susan Leavy (UCD) discussed AI’s societal impact, while Prof. Eamonn Keogh (UC Riverside) critiqued research in time series anomaly detection. AICS 2024 reaffirmed its role as Ireland’s leading AI research forum, fostering discussions on cutting-edge advancements and their societal implications. © 2024 Copyright for this paper by its authors.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Overview of the shared task on Fake News Detection from Social Media Text 3

Overview of the shared task on Fake News Detection from Soci...

引用

3rd Workshop on Speech and Language Technologies for Dravidian Languages, DravidianLangTech 2023

作者： Subramanian, Malliga Chakravarthi, Bharathi Raja Shanmugavadivel, Kogilavani Pandiyan, Santhiya Kumaresan, Prasanna Kumar Palani, Balasubramanian Singh, Muskaan Raja, Sandhiya Vanaja Mithunajha, S. Kongu Engineering College Tamil Nadu India Insight SFI Research Centre for Data Analytics School of Computer Science University of Galway Ireland Indian Institute of Information Technology Kottayam India Ulster University United Kingdom

ISBN: (纸本)9789544520854

The rapid proliferation of fake news has emerged as a significant challenge to the credibility of online information, largely due to the swift dissemination of content on social media platforms. This article provides a concise summary of the findings of the shared task on"Fake News Detection in Dravidian Languages1 - DravidianLangTech@RANLP 2023". The aim of this shared task is to categorize social media posts as either fake or original, specifically focusing on content in Malayalam. The shared task garnered participation from 8 teams who presented their systems. These systems encompassed a spectrum of methodologies including machine learning techniques and transformer-based models like MuRIL, XLMRoBERTa, and Indic BERT. Notably, the XLMRoBERTa-based model demonstrated exceptional performance, achieving a macro F1-score of 0.90. © DravidianLangTech 2023 - 3rd Workshop on Speech and Language Technologies for Dravidian Languages, associated with 14th International Conference on Recent Advances in Natural Language Processing, RANLP 2023 - Proceedings.

关键词： Learning systems

来源：评论

学校读者我要写书评

暂无评论

Integrated Energy Management System and Optimization in Smart Grids: Case of Smart Campus in Thailand

Integrated Energy Management System and Optimization in Smar...

引用

IEEE International Smart Cities Conference (ISC2)

作者： Nattapong Thanngam Sarawut Jansuwan Department of Business Analytics and Data Science Graduate School of Applied Statistics National Institute of Development Administration (NIDA) Bangkok Thailand

ISBN: (数字)9798350364316

ISBN: (纸本)9798350364323

Energy is vital for economic growth and powering daily activities. The primary aim of this work is to create an Energy Management System (EMS) by using an optimization method to effectively control energy sources, encompassing conventional power grids, solar cells, and batteries. The case study uses real-time power consumption data from NIDA, Thailand. The study compares and tests three energy management schemes: 1) the existing energy supply from the MEA grid (without a solar cell system), 2) the electricity supply from the MEA grid with the addition of a solar cell system, and 3) case 2 with the inclusion of the optimized EMS. We use time series forecasting combined with dynamic optimization to develop an efficient energy management plan. The findings suggest that implementing a solar cell system can result in annual savings for the university, and the optimized EMS can save an extra annually. The project, which incorporates a 100 kWh battery, which is economically viable, as it boasts a payback period of 5.61 years.

关键词： Power demand Smart cities Photovoltaic cells Time series analysis Power system dynamics Real-time systems Smart grids Batteries Energy management systems Forecasting

来源：评论

学校读者我要写书评

暂无评论

A Robust Score Test in G-computation for Covariate Adjustment in Randomized Clinical Trials Leveraging Different Variance Estimators via Influence Functions

arXiv

引用

arXiv 2025年

作者： Zhang, Xin Chu, Haitao Liu, Lin Roychoudhury, Satrajit Data Sciences and Analytics Pfizer Inc Division of Biostatistics and Health Data Science University of Minnesota United States Institute of Natural Sciences MOE-LSC School of Mathematical Sciences CMA-Shanghai SJTU-Yale Joint Center for Biostatistics and Data Science Shanghai Jiao Tong University China

G-computation has become a widely used robust method for estimating unconditional (marginal) treatment effects with covariate adjustment in the analysis of randomized clinical trials. Statistical inference in this context typically relies on the Wald test or Wald interval, which can be easily implemented using a consistent variance estimator. However, existing literature suggests that when sample sizes are small or when parameters of interest are near boundary values, Wald-based methods may be less reliable due to type I error rate inflation and insufficient interval coverage. In this article, we propose a robust score test for g-computation estimators in the context of two-sample treatment comparisons. The proposed test is asymptotically valid under simple and stratified (biased-coin) randomization schemes, even when regression models are misspecified. These test statistics can be conveniently computed using existing variance estimators, and the corresponding confidence intervals have closed-form expressions, making them convenient to implement. Through extensive simulations, we demonstrate the superior finite-sample performance of the proposed method. Finally, we apply the proposed method to reanalyze a completed randomized clinical trial. The new analysis using our proposed score test achieves statistical significance, whilst reducing the issue of type I error inflation. Copyright © 2025, The Authors. All rights reserved.

关键词： Regression analysis

来源：评论

学校读者我要写书评

暂无评论

Prediction of circularity of the holes in electrical discharge machining of inconel alloy using machine learning technique

引用

Interactions 2024年第1期245卷 1-10页

作者： Sujatha, M. Kavitha, B.C. Vimalarani, G. David, D. Beulah Muthukumar, K. Ram, S. Department of Electrical and Electronics Engineering Easwari Engineering College Chennai India Department of Electronics and Communication Engineering Hindustan Institute of Technology and Science Chennai India Department of Data Analytics Institute of Information Technology Saveetha School of Engineering Saveetha Institute of Medical and Technical Sciences Thandalam Tamilnadu Chennai India Department of Electrical and Electronics Engineering Sri Krishna College of Engineering and Technology Coimbatore India

Accuracy and finishing of the component is playing an essential role in metal forming process. The quality of the product is depending on accuracy of the shape and size. The measurement of holes is a challenging task in electric discharge machining (EDM) of materials. Sometimes, the diameter of the holes are different and its seems to be conical shape. The circularity of the hole is not maintained due to over current, low voltage and improper electrode materials. In this article, the circularity of the holes and its dimensions are analyzed by coordinate measuring machine (CMM). The experimental results are compared with predicted results with the aid of machine learning technique through regularized random forest (RRF) model. Inconel 718 alloy is used as a work material for this experimental investigation. The material characterization and its performance against the EDM process are studied. Prediction of circularity of the holes by machining of inconel alloy is perfectly done by machine learning technique. It’s used to enhance the accuracy of the hole and performance of the process. © The Author(s), under exclusive licence to Springer Nature Switzerland AG 2024.

关键词： Metal forming

来源：评论

学校读者我要写书评

暂无评论

MatchPrompt: Prompt-based Open Relation Extraction with Semantic Consistency Guided Clustering

MatchPrompt: Prompt-based Open Relation Extraction with Sema...

引用

2022 Conference on Empirical Methods in Natural Language Processing, EMNLP 2022

作者： Wang, Jiaxin Zhang, Lingling Liu, Jun Liang, Xi Zhong, Yujie Wu, Yaqiang Shaanxi Provincial Key Laboratory of Big Data Knowledge Engineering School of Computer Science and Technology Xi'an Jiaotong University China National Engineering Lab for Big Data Analytics Xi'an Jiaotong University China Lenovo Research Beijing China

Relation clustering is a general approach for open relation extraction (OpenRE). Current methods have two major problems. One is that their good performance relies on large amounts of labeled and pre-defined relational instances for pre-training, which are costly to acquire in reality. The other is that they only focus on learning a high-dimensional metric space to measure the similarity of novel relations and ignore the specific relational representations of clusters. In this work, we propose a new prompt-based framework named MatchPrompt, which can realize OpenRE with efficient knowledge transfer from only a few predefined relational instances as well as mine the specific meanings for cluster interpretability. To our best knowledge, we are the first to introduce a prompt-based framework for unlabeled clustering. Experimental results on different datasets show that MatchPrompt achieves the new SOTA results for OpenRE. © 2022 Association for Computational Linguistics.

关键词： Extraction

来源：评论

学校读者我要写书评

暂无评论

Investigating the Impact of Temporal Labeling of Emergency Department Visits for COVID-19: Comparing Healthcare Disparities Analyses Using Comprehensive, Single-Site data with National COVID Cohort Collaborative (N3C) data

Investigating the Impact of Temporal Labeling of Emergency D...

引用

2023 Systems and Information Engineering Design Symposium, SIEDS 2023

作者： Jones, Madeleine Winger, Aubrey Wernz, Christian Michel, Jonathan Jiang, Sihang Zhou, Andrea Hilton, Ebony Jade Zemmel, Mikhail Sengupta, Saurav Barnes, Kierah Loomba, Johanna Brown, Donald E. Wilcox, Adam B. Lee, Adam M. Graves, Alexis Anzalone, Alfred Manna, Amin Saha, Amit Olex, Amy Williams, Andrew E. Southerland, Andrew Girvin, Andrew T. Walden, Anita Sharathkumar, Anjali A. Amor, Benjamin Bates, Benjamin Hendricks, Brian Patel, Brijesh Alexander, Caleb Bramante, Carolyn Ward-Caviness, Cavin Madlock-Brown, Charisse Suver, Christine Chute, Christopher Dillon, Christopher Wu, Chunlei Schmitt, Clare Takemoto, Cliff Housman, Dan Gabriel, Davera Eichmann, David A. Mazzotti, Diego Boudreau, Eilis Hill, Elaine Zampino, Elizabeth Marti, Emily Carlson Pfaff, Emily R. French, Evan Koraishy, Farrukh M. Mariona, Federico Prior, Fred Sokos, George Martin, Greg Lehmann, Harold Spratt, Heidi Mehta, Hemalkumar Liu, Hongfang Sidky, Hythem Awori Hayanga, J.W. Pincavitch, Jami Clark, Jaylyn Richard Harper, Jeremy Islam, Jessica Ge, Jin Gagnier, Joel Saltz, Joel H. Buse, John Mathew, Jomol Rutter, Joni L. McMurry, Julie A. Guinney, Justin Starren, Justin Crowley, Karen Bradwell, Katie Rebecca Walters, Kellie M. Wilkins, Ken Gersing, Kenneth R. Cato, Kenrick Dwain Murray, Kimberly Kostka, Kristin Northington, Lavance Pyles, Lee Allan Misquitta, Leonie Cottrell, Lesley Portilla, Lili Deacy, Mariam Bissell, Mark M. Clark, Marshall Emmett, Mary Saltz, Mary Morrison Palchuk, Matvey B. Haendel, Melissa A. Adams, Meredith Temple-O'Connor, Meredith Kurilla, Michael G. Morris, Michele Qureshi, Nabeel Safdar, Nasia Garbarini, Nicole Sharafeldin, Noha Sadan, Ofer Francis, Patricia A. Burgoon, Penny Wung Robinson, Peter Payne, Philip R.O. Fuentes, Rafael Jawa, Randeep Erwin-Cohen, Rebecca Patel, Rena Moffitt, Richard A. Zhu, Richard L. Kamaleswaran, Rishi Hurley, Robert Miller, Robert T. Pyarajan, Saiju Michael, Sam G. Bozzette, Samuel Mallipattu, Sandeep Vedula, Satyanarayana Chapman, Scott O'Neil, Shawn T. Setoguchi, Soko Hong, Stephanie School of Data Science University of Virginia United States Data Science UVA Health United States University of Virginia Dept of Sys & Inf Eng United States University of Virginia ITHRIV United States Anesthesiology & CC Medicine UVA Health United States Research & Clinical Trial Analytics UVA Health United States UVA Health Department of Surgery United States

ISBN: (纸本)9798350300642

National COVID Cohort Collaborative (N3C) enclave provides health researchers with a rich dataset from 76 contributing clinical sites. However, the harmonized data lacks certain details available in sites' local electronic health records (EHRs), such as the principal diagnosis code for reported emergency department (ED) and inpatient (IP) visits. This means a principal diagnosis of COVID-19 can only be inferred by applying a time relationship between the visit dates and the record of infection and diagnosis. The purpose of this study is to perform a single-site sensitivity analysis modeled after an N3C study examining potential race-ethnicity based bias in hospitalization decisions during COVID-19 related ED visits. The analytic pipeline was first run in N3C, then reproduced locally with N3C data fields from a single-site, and finally run a third time using the additional principal diagnosis data. We find the effects of patient comorbidities and race-ethnicity groups on direct IP admittance to be consistent among the three cohorts with varying levels of statistical significance due to different sample sizes. © 2023 IEEE.

关键词： COVID-19

来源：评论

学校读者我要写书评

暂无评论

Multilingual Sentiment Analysis using Deep-Learning Architectures 5

Multilingual Sentiment Analysis using Deep-Learning Architec...

引用

5th International Conference on Smart Systems and Inventive Technology, ICSSIT 2023

作者： Dominic, Praveen Purushothaman, Niranjan Kumar, Anish Skanda Anil Prabagaran, A. Blessy, J. Angelin John, A. School of Graduate Studies MSc in Big Data Analytics Trent University Ontario Canada Visa Consolidated Support Services India Pvt. Ltd Bengaluru India School of Computing Science and Engineering Galgotias University Uttar Pradesh Greater Noida India

ISBN: (纸本)9781665474672

Machine learning techniques such as NLP (Natural language processing) play a key role in a context where mining social media data could add great value to governments of the world countries. The posts and tweets shared by the people on social media can be mined to infer the valuable 'mindset' of the people which is much required for any ruling government in the world. The objective of this study is to conduct sentiment analysis to mine the sentiment of the people regarding the ongoing war between Russia and Ukraine, using machine learning techniques. The idea is to analyze and infer if the countries have reacted in some way, considering the sentiment of their citizens, in the context of economic effects. The pipeline of the implementation associated starts with the data collection from social media such as Twitter and Reddit using Snscraper and the PRAW (Python Reddit API Wrapper). The larger posts from Reddit are handled by implementing suitable text summarization techniques. Sentiment analysis is performed for the social media data using the BERT transformer model. The non-English posts are translated to English using neural machine translation. Also, sentiment analysis is performed at various granularities such as the location and the people that are tagged using Named Entity Recognition techniques. Finally, a comparative analysis of the world countries' sentiment and their corresponding reliance on Russian oil is performed. © 2023 IEEE.

关键词： Sentiment analysis

来源：评论

学校读者我要写书评

暂无评论

Leaky Dropout: Balancing Performance and Overfitting

Leaky Dropout: Balancing Performance and Overfitting

引用

Artificial Intelligence and Knowledge Engineering (AIKE)

作者： Ryan Parker Ramazan Aygun School of Data Science and Analytics Kennesaw State University Kennesaw GA USA Department of Computer Science Kennesaw State University Kennesaw GA USA

ISBN: (数字)9798331517045

ISBN: (纸本)9798331517052

Dropout layers are widely used in neural networks for regularization, but can reduce performance in some cases. We present Leaky Dropout, a generalization of dropout that scales down a random selection of inputs at each layer by a hyperparameter-determined factor, rather than silencing them. Dropout, Gaussian Dropout and our novel Leaky Dropout were analyzed and compared using out-of-sample performance and with respect to overfitting. Leaky dropout outperformed the other two methods on out-of-sample data, with little to no overfitting, and outperformed baseline models with no dropout in most cases. The performance and training were inversely related to overfitting and dependent on the scaling factor and the dropout rate. On small training sets and complex models, Leaky Dropout served as a better sort of regularization than the other two methods.

关键词： Training Deep learning Resistance Knowledge engineering Neural networks Finance data models Neck Health and safety Overfitting

来源：评论

学校读者我要写书评

暂无评论

Evaluation and Mitigation of the Limitations of Large Language Models in Business Decision-Making

Evaluation and Mitigation of the Limitations of Large Langua...

引用

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

作者： Mohammad Al Khaldy Youcef Gheraibia Department of Business Intelligence and Data Analytics University of Petra Amman Jordan School of Computer Science and Informatics De Montfort University Leicester UK

ISBN: (数字)9798331523657

ISBN: (纸本)9798331523664

Large Language Models (LLMs) have emerged as transformative tools in business decision-making, offering capabilities such as automation, market analysis, and customer insights. However, their integration into business processes presents significant limitations, including biases in training data, lack of domain-specific knowledge, interpretability issues, and ethical concerns. These challenges can affect the quality and reliability of business decisions, potentially leading to biased or erroneous outcomes. This paper evaluates the key limitations of LLMs in business contexts and explores strategies to mitigate these issues. Proposed solutions include diversifying training data to reduce biases, fine-tuning models with industry-specific expertise, implementing explainable AI techniques for better transparency, and establishing ethical guidelines to safeguard against misuse. By addressing these limitations, businesses can harness the full potential of LLMs, ensuring more effective, fair, and accountable decision-making processes. This research provides a comprehensive approach for mitigating the risks associated with LLMs in business environments.

关键词： Industries Ethics Explainable AI Large language models Prevention and mitigation Decision making Training data Predictive analytics Business Guidelines

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：