检索结果-内蒙古大学图书馆

11th International Conference on Learning Representations, ICLR 2023

作者： Gosch, Lukas Sturm, Daniel Geisler, Simon Günnemann, Stephan Department of Computer Science Munich Data Science Institute Technical University of Munich Germany

Many works show that node-level predictions of Graph Neural Networks (GNNs) are unrobust to small, often termed adversarial, changes to the graph structure. However, because manual inspection of a graph is difficult, it is unclear if the studied perturbations always preserve a core assumption of adversarial examples: that of unchanged semantic content. To address this problem, we introduce a more principled notion of an adversarial graph, which is aware of semantic content change. Using Contextual Stochastic Block Models (CSBMs) and real-world graphs, our results uncover: i) for a majority of nodes the prevalent perturbation models include a large fraction of perturbed graphs violating the unchanged semantics assumption;ii) surprisingly, all assessed GNNs show over-robustness - that is robustness beyond the point of semantic change. We find this to be a complementary phenomenon to adversarial examples and show that including the label-structure of the training graph into the inference process of GNNs significantly reduces over-robustness, while having a positive effect on test accuracy and adversarial robustness. Theoretically, leveraging our new semantics-aware notion of robustness, we prove that there is no robustness-accuracy tradeoff for inductively classifying a newly added node. © 2023 11th International Conference on Learning Representations, ICLR 2023. All rights reserved.

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Enhanced Multimodal Hate Video Detection via Channel-wise and Modality-wise Fusion 24

Enhanced Multimodal Hate Video Detection via Channel-wise an...

引用

24th IEEE International Conference on data Mining Workshops, ICDMW 2024

作者： Zhang, Yinghui Chen, Tailin Zhang, Yuchen Fu, Zeyu University of Exeter Department of Computer Science United Kingdom University of Essex Institute for Analytics and Data Science United Kingdom

ISBN: (纸本)9798331530631

The rapid rise of video content on platforms such as TikTok and YouTube has transformed information dissemination, but it has also facilitated the spread of harmful content, particularly hate videos. Despite significant efforts to combat hate speech, detecting these videos remains challenging due to their often implicit nature. Current detection methods primarily rely on unimodal approaches, which inadequately capture the complementary features across different modalities. While multimodal techniques offer a broader perspective, many fail to effectively integrate temporal dynamics and modality-wise interactions essential for identifying nuanced hate content. In this paper, we present CMFusion, an enhanced multimodal hate video detection model utilizing a novel Channel-wise and Modality-wise Fusion Mechanism. CMFusion first extracts features from text, audio, and video modalities using pre-trained models and then incorporates a temporal cross-attention mechanism to capture dependencies between video and audio streams. The learned features are then processed by channel-wise and modality-wise fusion modules to obtain informative representations of videos. Our extensive experiments on a real-world dataset demonstrate that CMFusion significantly outperforms five widely used baselines in terms of accuracy, precision, recall, and F1 score. Comprehensive ablation studies and parameter analyses further validate our design choices, highlighting the model's effectiveness in detecting hate videos. The source codes will be made publicly available at https://***/EvelynZ10/cmfusion. © 2024 IEEE.

关键词： Hate Video Detection Multimodal Fusion Temporal Cross-Attention

来源：评论

学校读者我要写书评

暂无评论

Robust Inference for Censored Quantile Regression

引用

Journal of Systems science & Complexity 2024年

作者： TANG Yuanyuan WANG Xiaorui ZHU Jianming LIN Hongmei TANG Yanlin TONG Tiejun School of Statistics and Information Shanghai University of International Business and Economics Department of Statistics and Data Science Southern University of Science and Technology School of Statistics KLATASDS-MOE East China Normal University Department of Mathematics Hong Kong Baptist University

In various fields such as medical science and finance, it is not uncommon that the data are heavy-tailed and/or not fully observed, calling for robust inference methods that can deal with the outliers and incompleteness efficiently. In this paper, the authors propose a rank score test for quantile regression with fixed censored responses, based on the standard quantile regression in an informative subset which is computationally efficient and robust. The authors further select the informative subset by the multiply robust propensity scores, and then derive the asymptotic properties of the proposed test statistic under both the null and local alternatives. Moreover, the authors conduct extensive simulations to verify the validity of the proposed test, and apply it to a human immunodeficiency virus data set to identify the important predictors for the conditional quantiles of the censored viral load.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Correlation between first trimester placental growth factor levels and skin microvascular reactivity assessed by laser speckle contrast imaging - a cross-sectional study

引用

Placenta 2025年 167卷 187-192页

作者： Huri, Mor Abati, Isabella Bartolini, Chiara Piacenza, Alessia Tofani, Lorenzo Vallario, Arianna Di Tommaso, Mariarosaria Seravalli, Viola Department of Health Sciences Obstetrics and Gynaecology Unit University of Florence Florence Italy Department of Statistics Computer Science Applications University of Florence Florence Italy

Objectives: To investigate the association between first-trimester placental growth factor (PlGF) levels and maternal skin microvascular reactivity, as assessed by laser speckle contrast imaging (LSCI) combined with post-occlusive reactive hyperemia. Additionally, to explore the correlations between maternal microvascular function and other first-trimester serum biochemical and biophysical markers. Methods: Fifty-three patients carrying a singleton gestation were enrolled during their routine first trimester scan. Skin blood flux at the dorsal hand was recorded using LSCI before, during, and after a 3-min arterial occlusion. Microvascular reactivity parameters were calculated and compared with maternal serum biochemical markers (PlGF, pregnancy-associated plasma protein A [PAPP-A], and free beta-human chorionic gonadotropin [free β-hCG]), expressed as multiples of the median, and with maternal biophysical markers. Results: PlGF levels showed a moderate positive correlation with base-to-peak flux (r = 0.51, 95 % confidence interval, CI, 0.27–0.69) and a weak but statistically significant positive correlation with peak flux (r = 0.31, 95 % CI 0.04–0.59). PAPP-A levels above the median were associated with higher base-to peak flux compared to PAPP-A below the median (253.41 % versus 215.08 %, p = 0.02). A moderate positive correlation was also found between free β-hCG and peak flux (r = 0.4, 95 % CI 0.15–0.60). No correlations were found between the parameters of hyperemic response and maternal biophysical markers. Conclusions: Maternal first-trimester skin microvascular reactivity indices correlate positively with serum placental biomarker levels, particularly PlGF. This suggests that maternal peripheral microvascular function, assessed by LSCI, may reflect placental microcirculation. Further studies are warranted to determine whether this tool could serve as an early marker of placental function. © 2025 The Authors

关键词： Free beta-human chorionic gonadotropin Laser-speckle-contrast-imaging Microvascular reactivity Placental-growth-factor Preeclampsia Pregnancy Pregnancy-associated plasma protein A

来源：评论

学校读者我要写书评

暂无评论

SAMPLING-FREE INFERENCE FOR AB-INITIO POTENTIAL ENERGY SURFACE NETWORKS 11

SAMPLING-FREE INFERENCE FOR AB-INITIO POTENTIAL ENERGY SURFA...

引用

11th International Conference on Learning Representations, ICLR 2023

作者： Gao, Nicholas Günnemann, Stephan Department of Computer Science Munich Data Science Institute Technical University of Munich Germany

Recently, it has been shown that neural networks not only approximate the ground-state wave functions of a single molecular system well but can also generalize to multiple geometries. While such generalization significantly speeds up training, each energy evaluation still requires Monte Carlo integration which limits the evaluation to a few geometries. In this work, we address the inference shortcomings by proposing the Potential learning from ab-initio Networks (PlaNet) framework, in which we simultaneously train a surrogate model in addition to the neural wave function. At inference time, the surrogate avoids expensive Monte-Carlo integration by directly estimating the energy, accelerating the process from hours to milliseconds. In this way, we can accurately model high-resolution multi-dimensional energy surfaces for larger systems that previously were unobtainable via neural wave functions. Finally, we explore an additional inductive bias by introducing physically-motivated restricted neural wave function models. We implement such a function with several additional improvements in the new PESNet++ model. In our experimental evaluation, PlaNet accelerates inference by 7 orders of magnitude for larger molecules like ethanol while preserving accuracy. Compared to previous energy surface networks, PESNet++ reduces energy errors by up to 74 %. © 2023 11th International Conference on Learning Representations, ICLR 2023. All rights reserved.

关键词： Wave functions

来源：评论

学校读者我要写书评

暂无评论

UNDERSTANDING THE GENERALIZATION OF ADAM IN LEARNING NEURAL NETWORKS WITH PROPER REGULARIZATION 11

UNDERSTANDING THE GENERALIZATION OF ADAM IN LEARNING NEURAL ...

引用

11th International Conference on Learning Representations, ICLR 2023

作者： Zou, Difan Cao, Yuan Li, Yuanzhi Gu, Quanquan Department of Computer Science Institute of Data Science The University of Hong Kong Hong Kong Department of Statistics & Actuarial Science The University of Hong Kong Hong Kong Machine Learning Department Carnegie Mellon University United States Department of Computer Science University of California Los Angeles United States

Adaptive gradient methods such as Adam have gained increasing popularity in deep learning optimization. However, it has been observed in many deep learning applications such as image classification, Adam can converge to a different solution with a worse test error compared to (stochastic) gradient descent, even with a fine-tuned regularization. In this paper, we provide a theoretical explanation for this phenomenon: we show that in the nonconvex setting of learning over-parameterized two-layer convolutional neural networks starting from the same random initialization, for a class of data distributions (inspired from image data), Adam and gradient descent (GD) can converge to different global solutions of the training objective with provably different generalization errors, even with weight decay regularization. In contrast, we show that if the training objective is convex, and the weight decay regularization is employed, any optimization algorithms including Adam and GD will converge to the same solution if the training is successful. This suggests that the generalization gap between Adam and SGD in the presence of weight decay regularization is closely tied to the nonconvex landscape of deep learning optimization, which cannot be covered by the recent neural tangent kernel (NTK) based analysis. © 2023 11th International Conference on Learning Representations, ICLR 2023. All rights reserved.

关键词： Gradient methods

来源：评论

学校读者我要写书评

暂无评论

A Plastic Waste Hotspot Detection System and Its Application to Measuring Environmental Improvement 1

A Plastic Waste Hotspot Detection System and Its Application...

引用

1st IEEE International Symposium on Consumer Technology, ISCT 2024

作者： Sasaki, Shiori Nakamura, Ryota Wibowo, Prasetyo Musashino University Department of Data Science Tokyo Japan Musashino University Faculty of Data Science Tokyo Japan Politeknik Elektronika Negeri Department of Informatics and Computer Engineering Surabaya Indonesia

ISBN: (纸本)9798350365191

This paper describes a Plastic Waste Hotspot Detection System which has been developed in an international collaborative research project to realize an 'Environmental AI-Human Actions integration' with marine plastic waste reduction actions. In this project, various local conditions are obtained to the system to realize an environmentally resilient society. The nature conservation activities such as marine plastic waste collection and reduction actions are implemented based on these conditions. In the process of integration of Environmental AI and Human Actions, the effects of these activities and activists' contribution to environmental improvement and global-scale knowledge are accumulated, shared, analyzed, and visualized through global-scale analysis. This paper describes the Plastic Waste Hotspot Detection system as one of the systems that integrate AI systems and human actions. This paper examines the feasibility and effectiveness of the system through the fieldwork tests in Phuket, Thailand and Surabaya, Indonesia. © 2024 IEEE.

关键词： Plastic recycling

来源：评论

学校读者我要写书评

暂无评论

DOS: DIVERSE OUTLIER SAMPLING FOR OUT-OF-DISTRIBUTION DETECTION 12

DOS: DIVERSE OUTLIER SAMPLING FOR OUT-OF-DISTRIBUTION DETECT...

引用

12th International Conference on Learning Representations, ICLR 2024

作者： Jiang, Wenyu Cheng, Hao Chen, Mingcai Wang, Chongjun Wei, Hongxin Department of Statistics and Data Science Southern University of Science and Technology China State Key Laboratory for Novel Software Technology Nanjing University China

Modern neural networks are known to give overconfident predictions for out-of-distribution inputs when deployed in the open world. It is common practice to leverage a surrogate outlier dataset to regularize the model during training, and recent studies emphasize the role of uncertainty in designing the sampling strategy for outlier datasets. However, the OOD samples selected solely based on predictive uncertainty can be biased towards certain types, which may fail to capture the full outlier distribution. In this work, we empirically show that diversity is critical in sampling outliers for OOD detection performance. Motivated by the observation, we propose a straightforward and novel sampling strategy named DOS (Diverse Outlier Sampling) to select diverse and informative outliers. Specifically, we cluster the normalized features at each iteration, and the most informative outlier from each cluster is selected for model training with absent category loss. With DOS, the sampled outliers efficiently shape a globally compact decision boundary between ID and OOD data. Extensive experiments demonstrate the superiority of DOS, reducing the average FPR95 by up to 25.79% on CIFAR-100 with TI-300K. © 2024 12th International Conference on Learning Representations, ICLR 2024. All rights reserved.

关键词： statistics

来源：评论

学校读者我要写书评

暂无评论

On the analytical soliton approximations to fractional forced Korteweg-de Vries equation arising in fluids and plasmas using two novel techniques

引用

Communications in Theoretical Physics 2024年第8期76卷 1-10页

作者： Weaam Alhejaili Emad A Az-Zo’bi Rasool Shah S A El-Tantawy Department of Mathematical Sciences College of SciencePrincess Nourah bint Abdulrahman UniversityPO Box 84428Riyadh 11671Saudi Arabia Department Mathematics and Statistics Mutah UniversityAl-KarakJordan Department of Computer Science and Mathematics Lebanese American UniversityBeirutLebanon Department of Physics Faculty of SciencePort Said UniversityPort Said 42521Egypt Department of Physics Faculty of ScienceAl-Baha UniversityAl-Baha P.O.Box 1988Saudi Arabia

The current investigation examines the fractional forced Korteweg-de Vries(FF-KdV) equation,a critically significant evolution equation in various nonlinear branches of science. The equation in question and other associated equations are widely acknowledged for their broad applicability and potential for simulating a wide range of nonlinear phenomena in fluid physics, plasma physics, and various scientific domains. Consequently, the main goal of this study is to use the Yang homotopy perturbation method and the Yang transform decomposition method, along with the Caputo operator for analyzing the FF-KdV equation. The derived approximations are numerically examined and discussed. Our study will show that the two suggested methods are helpful, easy to use, and essential for looking at different nonlinear models that affect complex processes.

关键词： forced fractional KdV equation Caputo operator Yang homotopy perturbation method Yang transform decomposition method solitons

来源：评论

学校读者我要写书评

暂无评论

Monitoring photochemical pollutants based on symbolic interval-valued data analysis

引用

Advances in data Analysis and Classification 2023年第4期17卷 897-926页

作者： Lin, Liang-Ching Guo, Meihui Lee, Sangyeol Department of Statistics and Institute of Data Science National Cheng Kung University Tainan Taiwan Department of Applied Mathematics National Sun Yat-sen University Kaohsiung Taiwan Department of Statistics Seoul National University Seoul08826 Korea Republic of

This study considers monitoring photochemical pollutants for anomaly detection based on symbolic interval-valued data analysis. For this task, we construct control charts based on the principal component scores of symbolic interval-valued data. Herein, the symbolic interval-valued data are assumed to follow a normal distribution, and an approximate expectation formula of order statistics from the normal distribution is used in the univariate case to estimate the mean and variance via the method of moments. In addition, we consider the bivariate case wherein we use the maximum likelihood estimator calculated from the likelihood function derived under a bivariate copula. We also establish the procedures for the statistical control chart based on the univariate and bivariate interval-valued variables, and the procedures are potentially extendable to higher dimensional cases. Monte Carlo simulations and real data analysis using photochemical pollutants confirm the validity of the proposed method. The results particularly show the superiority over the conventional method that uses the averages to identify the date on which the abnormal maximum occurred. © 2022, Springer-Verlag GmbH Germany, part of Springer Nature.

关键词： Control charts

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：