检索结果-内蒙古大学图书馆

Assessing high-order effects in feature importance via predictability decomposition

Physical Review E 2025年第3期111卷 L033301-L033301页

作者： Marlis Ontivero-Ortega Luca Faes Jesus M. Cortes Daniele Marinazzo Sebastiano Stramaglia Dipartimento Interateneo di Fisica Università degli Studi di Bari Aldo Moro and INFN Sezione di Bari 70126 Bari Italy Dipartimento di Ingegneria Università di Palermo 90128 Palermo Italy Faculty of Technical Sciences University of Novi Sad 21000 Novi Sad Serbia Biocruces-Bizkaia Health Research Institute 48903 Barakaldo Spain Biomedical Research Doctorate Program University of the Basque Country 48940 Leioa Spain Department of Cell Biology and Histology University of the Basque Country 48940 Leioa Spain IKERBASQUE Basque Foundation for Science 48009 Bilbao Spain Department of Data Analysis Ghent University 9000 Ghent Belgium

Building on recent advances in describing redundancy and synergy in multivariate interactions among random variables, we propose an approach to quantify cooperative effects in feature importance, a key technique in explainable artificial intelligence. Specifically, we introduce an adaptive version of the widely used metric Leave One Covariate Out (LOCO), designed to disentangle high-order effects involving a particular input feature in regression problems. LOCO measures the reduction in prediction error when the feature of interest is added to the set of features used in regression. Unlike the standard approach that computes LOCO using all available features, our method identifies the subsets of features that maximize and minimize LOCO. This results in a decomposition of LOCO into a two-body component and higher-order components (redundant and synergistic), while also identifying the features that contribute to these high-order effects in conjunction with the driving feature. We demonstrate the effectiveness of the proposed method in a benchmark dataset related to wine quality and to proton versus pion discrimination using simulated detector measurements generated by GEANT.

关键词： Regression analysis

来源：评论

学校读者我要写书评

暂无评论

Importance of Training Sample Based on SRGAN

Importance of Training Sample Based on SRGAN

引用

2021 IEEE International Conference on Computer science, Electronic Information Engineering and Intelligent Control Technology, CEI 2021

作者： An, Zimu Jin, Haici Zhou, Xinrui University of California San Diego Halicioǧlu Data Science Institute San DiegoCA United States Peking University Yuanpei College Beijing China The Hong Kong University of Science and Technology Interdisciplinary Program Office Hong Kong

ISBN: (纸本)9780738146492

Generative Adversarial Network has high capabilities for generating realistic images and other real-life applications due to its generator-discriminator interactions. Like other neural networks in deep learning, an efficient training process is indispensable to model performance. With this feature, many prior works that aimed to optimize the learning process within this structure focused on importance sampling and training datasets. Aside from finding important data and accelerating the learning process, these studies have a common assumption: gradient descent can directly indicate data importance. However, for image data in high dimensional space, gradient descent could potentially be misleading. Hence, this study offers a different insight into data importance: we want to verify if using a gradient descent path is a reliable method for indicating data importance. Furthermore, if such important data exists, we also want to know if the importance persists in multiple training processes (with different input orders). After conducting five experiments with each trial rearranging the input order of the training set on the SRGAN model, we found that data importance cannot be indicated by any individual metric such as gradient, and the order has a great impact on the gradient descent path. © 2021 IEEE.

关键词： Importance sampling

来源：评论

学校读者我要写书评

暂无评论

EUDRL: Explainable Uncertainty-Based Deep Reinforcement Learning for Portfolio Management

EUDRL: Explainable Uncertainty-Based Deep Reinforcement Lear...

引用

International Conference on Computing and Networking Technology (ICCNT)

作者： Jia Lina Shahab S. Banda Hao-Ting Paib Bharat S. Rawal International Graduate School of Artificial Intelligence Douliu City Yunlin Taiwan Bachelor Program of Big Data Applications National Pingtung University Pingtung Taiwan Department of Computer Science Digital Technologies University Ggrambling La USA

ISBN: (数字)9798350370249

ISBN: (纸本)9798350370270

Supervised deep learning (SDL) has shown remarkable success in various financial applications, such as stock prediction and fraud detection. However, SDL’s reliance on class labels renders it unsuitable for portfolio management (PM) tasks, where such labels are often unavailable. To address this limitation, we propose a novel two-level architecture based on deep reinforcement learning (DRL) for PM, which does not require class labels. Our approach comprises several local agents that provide trading decisions and uncertainty assessments for individual stocks, and a global agent that makes portfolio management decisions based on the outputs of the local agents. Additionally, we incorporate the concept of explainable AI (XAI) into our framework using the SHAP (Shapley additive explanations) method, enhancing the transparency and interpretability of the global agent’s decisions. Our experimental results demonstrate that the proposed architecture consistently yields profitable outcomes in the market.

关键词： Uncertainty Correlation Additives Explainable AI Computational modeling Computer architecture Deep reinforcement learning Fraud Portfolios

来源：评论

学校读者我要写书评

暂无评论

Twitter-based text classification using svm for weather information system 6

Twitter-based text classification using svm for weather info...

引用

6th International Conference on Information Management and Technology, ICIMTech 2021

作者： Purwandari, Kartika Rahutomo, Reza Sigalingging, Join W. C. Ajy Kusuma, Mahisa Prasetyo, Aji Pardamean, Bens Research Center Bina Nusantara University Bioinformatics and Data Science Jakarta11480 Indonesia BMKG Meteorological Climatological and Geophysical Agency Database Center Division Jakarta10720 Indonesia Bina Nusantara University BINUS Graduate Program - Master of Computer Science Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9781665449373

The popularity of social media amplified the amount of text data that is used to enrich text classification research with machine learning approach. The modernization in weather information system is needed to produce real-time weather information using the involvement of Twitter users. GetOldTweets was used to cover data acquisition from Twitter. Support Vector Machine (SVM) was deployed as a machine learning model for text classification. The result was delivered in Esri Maps features. It facilitated a social sensing for Indonesia weather information system by using Twitter posts from Indonesian users with 93% accuracy. © 2021 IEEE.

关键词： data acquisition

来源：评论

学校读者我要写书评

暂无评论

The KIND dataset: A Social Collaboration Approach for Nuanced Dialect data Collection 18

The KIND Dataset: A Social Collaboration Approach for Nuance...

引用

18th Conference of the European Chapter of the Association for Computational Linguistics, EACL 2024 - Student Research Workshop, SRW 2024

作者： Yamani, Asma Z. Alziyady, Raghad AlYami, Reem Albelali, Salma A. Abouhagar, Leina Almulhim, Jawharah Alsulami, Amjad Alfarraj, Motaz Al-Zaidy, Rabeah Department of Information and Computer Science King Fahd University of Petroleum & Minerals Saudi Arabia Center for Integrative Petroleum Research CIPR King Fahd University of Petroleum & Minerals Saudi Arabia Department of Electrical Engineering King Fahd University of Petroleum & Minerals Saudi Arabia SDAIA-KFUPM Joint Research Center for AI King Fahd University of Petroleum & Minerals Saudi Arabia Preparatory Year Program King Fahd University of Petroleum & Minerals Saudi Arabia Department of Computer Science Imam Abdulrahman Bin Faisal University Saudi Arabia Saudi Data & AI Authority Saudi Arabia

ISBN: (纸本)9798891760905

Nuanced dialects are a linguistic variant that pose several challenges for NLP models and techniques. One of the main challenges is the limited amount of datasets to enable extensive research and experimentation. We propose an approach for efficiently collecting nuanced dialectal datasets that are not only of high quality, but are versatile enough to be multipurpose as well. To test our approach we collect the KIND corpus, which is a collection of fine-grained Arabic dialect data. The data is short texts, and unlike many nuanced dialectal datasets, it is curated manually through social collaboration efforts as opposed to being crawled from social media. The collaborative approach is incentivized through educational gamification and competitions for which the community itself benefits from the open source dataset. Our approach aims to achieve: (1) coverage of dialects from under-represented groups and fine-grained dialectal varieties, (2) provide aligned parallel corpora for translation between Modern Standard Arabic (MSA) and multiple dialects to enable translation and comparison studies, (3) promote innovative approaches for nuanced dialect data collection. We explain the steps for the competition as well as the resulting datasets and the competing data collection systems. The KIND dataset is shared with the research community. © 2024 Association for Computational Linguistics.

关键词： data acquisition

来源：评论

学校读者我要写书评

暂无评论

An equivalent condition for abelian varieties over finite fields to have QM

arXiv

引用

arXiv 2023年

作者： Arai, Keisuke Takai, Yuuki Department of Mathematics School of Science and Technology for Future Life Tokyo Denki University 5 Senju Asahi-cho Adachi-ku Tokyo120-8551 Japan Mathematics Science Data Science and AI Program Academic Foundations Programs Kanazawa Institute of Technology 7-1 Ohgigaoka Ishikawa Nonoichi921-8501 Japan

In this paper, we give an equivalent condition for an abelian variety over a finite field to have multiplication by a quaternion algebra over a number field. We prove the result by combining Tate’s classification of the endomorphism algebras of abelian varieties over finite fields with Yu’s criterion of the existence of homomorphisms between semi-simple *** Codes 11G10 (Primary) 11R52, 14K05 (Secondary) Copyright © 2023, The Authors. All rights reserved.

关键词： Algebra

来源：评论

学校读者我要写书评

暂无评论

Deep-learning triage of 3D pathology data for improved disease detection while reducing pathologist workloads

Deep-learning triage of 3D pathology data for improved disea...

引用

Microscopy Histopathology and Analytics, Microscopy 2024 - Part of Optica Biophotonics Congress: Biomedical Optics

作者： Gao, Gan Wang, Fiona Brenes, David Song, Andrew H. Chow, Sarah S.L. Mahmood, Faisal Liu, Jonathan T.C. Department of Mechanical Engineering University of Washington SeattleWA98195 United States Department of Computer Science University of Washington SeattleWA98195 United States Department of Pathology Brigham and Women's Hospital Harvard Medical School BostonMA02115 United States Department of Pathology Massachusetts General Hospital Harvard Medical School BostonMA02114 United States Cancer Program Broad Institute of Harvard and MIT CambridgeMA02142 United States Data Science Program Dana-Farber Cancer Institute BostonMA02215 United States Department of Bioengineering University of Washington SeattleWA98195 United States Department of Laboratory Medicine & Pathology University of Washington School of Medicine SeattleWA98195 United States

3D pathology can potentially improve disease detection, but the datasets are too large to review. We're developing a deep-learning-based triage method to identify the highest-risk 2D sections within 3D pathology d... 详细信息

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

PROCONSUL: PRObabilistic exploration of CONnectivity Significance patterns for disease modULe discovery

PROCONSUL: PRObabilistic exploration of CONnectivity Signifi...

引用

2022 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2022

作者： Luca, Riccardo De Carfora, Marco Blanco, Gonzalo Mastropietro, Andrea Petti, Manuela Tieri, Paolo Sapienza University of Rome Data Science Program Rome Italy Sapienza University of Rome DIAG Dept. Comp. Contr. Manag. Eng. Rome Italy IAC Institute for Applied Computing CNR National Research Council Rome Italy

ISBN: (纸本)9781665468190

The possibility to computationally prioritize candidate disease genes capitalizing on existing information has led to a speedup in the discovery of new methods. Many gene discovery techniques exploit network data, like protein-protein interactions (PPIs), in order to extract knowledge from the network structure relying on several network metrics. We here present PROCONSUL, a method that builds on top of the concept of connectivity significance (CS) and exploits the idea of probabilistic exploration of the space of putative disease genes. We show that our methodology is able to outperform the state-of-the-art tool based on CS in several settings, and propose different, effective gene discovery strategies according to specific disease network properties. © 2022 IEEE.

关键词： Proteins

来源：评论

学校读者我要写书评

暂无评论

College Spread of COVID-19 in Ohio

College Spread of COVID-19 in Ohio

引用

International Conference on Computational science and Computational Intelligence (CSCI)

作者： Akinkunle Akinola Akoh Atadoga Kimberlyn Brooks Vibhuti Chandna Robert C. Green Data Science Program Bowling Green State University Bowling Green OH USA Dept. of Computer Science Bowling Green State University Bowling Green OH USA

To determine whether Ohio college re-opening plans were effective in controlling the spread of COVID-19, cumulative case counts by county were gathered to compare various metrics related to the spread of COVID-19 cases between counties with NCAA colleges and counties without NCAA colleges. Various non-parametric statistical tests were used to determine if the samples were similar, and the analysis found the differences were statistically significant. Metropolitan and non-metropolitan groupings were also added to further subdivide the data set, but the analysis found no statistically significant differences in this case.

关键词：

来源：评论

学校读者我要写书评

暂无评论

SARA: Semantic-assisted Reinforced Active Learning for Entity Alignment

SARA: Semantic-assisted Reinforced Active Learning for Entit...

引用

International Joint Conference on Neural Networks (IJCNN)

作者： Ching-Hsuan Liu Chih-Ming Chen Jing-Kai Lou Ming-Feng Tsai Jiun-Lang Huang Chuan-Ju Wang Data Science Degree Program National Taiwan University Academia Sinica Taiwan Research Center for Information Technology Innovation Academia Sinica KKCompany Taiwan Department of Computer Science National Chengchi University Taiwan Department of Electrical Engineering National Taiwan University Taiwan Research Center for Information Technology Innovation Academia Sinica Taiwan

ISBN: (数字)9798350359312

ISBN: (纸本)9798350359329

This paper introduces SARA, a semantic-assisted reinforced active learning framework for enhancing entity alignment (EA) under limited supervision scenarios. SARA addresses the challenges of EA in real-world scenarios, including knowledge graph heterogeneity and limited training ground truth. SARA effectively selects valuable entity pairs with limited labeled data by combining reinforced active learning and semantic information. It utilizes a pair-wise language model based on Sentence-BERT to learn informative name embeddings that capture entity name semantics. These embeddings are combined with structural embeddings and trained using a novel semantic-assisted alignment loss. Extensive experiments on benchmark datasets and a real-world dataset demonstrate the superiority of SARA over existing approaches, particularly in limited labeled data scenarios. The paper also provides insights into fine-tuning strategies, presents ablation studies, and conducts sensitivity analyses to validate the effectiveness of SARA.

关键词： Training Sensitivity analysis Semantics Neural networks Knowledge graphs Benchmark testing data models

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：