检索结果-内蒙古大学图书馆

BIM: Improving Graph Neural Networks with Balanced Influence Maximization

学校读者我要写书评

暂无评论

BIM: Improving Graph Neural Networks with Balanced Influence...

International Conference on data Engineering

作者： Wentao Zhang Xinyi Gao Ling Yang Meng Cao Ping Huang Jiulong Shan Hongzhi Yin Bin Cui Center for Machine Learning Research Peking University Institute of Advanced Algorithms Research Shanghai National Engineering Labratory for Big Data Analytics and Applications The University of Queensland Australia Key Lab of High Confidence Software Technologies Peking University Apple Inc. Institute of Computational Social Science Peking University Qingdao

ISBN: (数字)9798350317152

ISBN: (纸本)9798350317169

The imbalanced data classification problem has aroused lots of concerns from both academia and industry since data imbalance is a widespread phenomenon in many real-world scenarios. Although this problem has been well researched from the view of imbalanced class samples, we further argue that graph neural networks (GNNs) expose a unique source of imbalance from the influenced nodes of different classes of labeled nodes, i.e., labeled nodes are imbalanced in terms of the number of nodes they influenced during the influence propagation in GNNs. To tackle this previously unexplored influence-imbalance issue, we connect social influence maximization with the imbalanced node classification problem and propose balanced influence maximization (BIM). Specifically, BIM greedily assigns the pseudo label to the node which can maximize the number of influenced nodes in GNN training while making the influence of each class more balance. Experimental results on five public datasets demonstrate the effectiveness of our method in relieving the influence-imbalance issue. For example, when training a GCN with an imbalance ratio of 0.1, BIM significantly outperforms the most competitive baseline by 0.6% -9.8% in five public datasets in terms of the F1 score.

关键词： Training Industries data engineering Graph neural networks

Enhancing Predictive Imaging Biomarker Discovery Through Treatment Effect analysis

学校读者我要写书评

暂无评论

Enhancing Predictive Imaging Biomarker Discovery Through Tre...

IEEE Workshop on Applications of Computer Vision (WACV)

作者： Shuhan Xiao Lukas Klein Jens Petersen Philipp Vollmuth Paul F. Jaeger Klaus H. Maier-Hein German Cancer Research Center (DKFZ) Heidelberg Division of Medical Image Computing Germany Faculty of Mathematics and Computer Science Heidelberg University Germany DKFZ Heidelberg Interactive Machine Learning Group Germany Institute for Machine Learning ETH Zürich Switzerland DKFZ Heidelberg Helmholtz Imaging Germany Division for Computational Radiology Clinical AI (CCIBonn.ai) Clinic for Neuroradiology University Hospital Bonn Germany Medical Faculty Bonn University of Bonn Germany Department of Radiation Oncology Pattern Analysis and Learning Group Heidelberg University Hospital Germany

ISBN: (数字)9798331510831

ISBN: (纸本)9798331510848

Identifying predictive covariates, which forecast individual treatment effectiveness, is crucial for decision-making across different disciplines such as personalized medicine. These covariates, referred to as biomarkers, are extracted from pretreatment data, often within randomized controlled trials, and should be distinguished from prognostic biomarkers, which are independent of treatment assignment. Our study focuses on discovering predictive imaging biomarkers, specific image features, by leveraging pretreatment images to uncover new causal relationships. Unlike laborintensive approaches relying on handcrafted features prone to bias, we present a novel task of directly learning predictive features from images. We propose an evaluation protocol to assess a model's ability to identify predictive imaging biomarkers and differentiate them from purely prognostic ones by employing statistical testing and a comprehensive analysis of image feature attribution. We explore the suitability of deep learning models originally developed for estimating the conditional average treatment effect (CATE) for this task, which have been assessed primarily for their precision of CATE estimation while overlooking the evaluation of imaging biomarker discovery. Our proof-of-concept analysis demonstrates the feasibility and potential of our approach in discovering and validating predictive imaging biomarkers from synthetic outcomes and real-world image datasets. Our code is available at https://***/MIC-DKFZ/predictive_image_biomarker_analysis.

关键词： Analytical models Statistical analysis Biological system modeling Precision medicine Decision making Imaging Estimation Biomarkers Predictive models Feature extraction

Improving generative model-based unfolding with Schrödinger bridges

学校读者我要写书评

暂无评论

Physical Review D 2024年第7期109卷 076011-076011页

作者： Sascha Diefenbacher Guan-Horng Liu Vinicius Mikuni Benjamin Nachman Weili Nie Physics Division Lawrence Berkeley National Laboratory Berkeley California 94720 USA Autonomous Control and Decision Systems Laboratory Georgia Institute of Technology Atlanta Georgia 30332 USA National Energy Research Scientific Computing Center Berkeley Lab Berkeley California 94720 USA Berkeley Institute for Data Science University of California Berkeley California 94720 USA Machine Learning Research Group NVIDIA Research

machine learning-based unfolding has enabled unbinned and high-dimensional differential cross section measurements. Two main approaches have emerged in this research area; one based on discriminative models and one based on generative models. The main advantage of discriminative models is that they learn a small correction to a starting simulation while generative models scale better to regions of phase space with little data. We propose to use Schrödinger bridges and diffusion models to create sbunfold, an unfolding approach that combines the strengths of both discriminative and generative models. The key feature of sbunfold is that its generative model maps one set of events into another without having to go through a known probability density as is the case for normalizing flows and standard diffusion models. We show that sbunfold achieves excellent performance compared to state of the art methods on a synthetic Z+jets dataset.

关键词： machine learning Particle detectors Probability theory

Double variance reduction: a smoothing trick for composite optimization problems without first-order gradient 24

学校读者我要写书评

暂无评论

Double variance reduction: a smoothing trick for composite o...

Proceedings of the 41st International Conference on machine learning

作者： Hao Di Haishan Ye Yueling Zhang Xiangyu Chang Guang Dai Ivor W. Tsang Center for Intelligent Decision-Making and Machine Learning School of Management Xi'an Jiaotong University China and SGIT AI Lab State Grid Corporation of China International Business School Beijing Foreign Studies University Beijing China Center for Intelligent Decision-Making and Machine Learning School of Management Xi'an Jiaotong University China SGIT AI Lab State Grid Corporation of China CFAR and IHPC Agency for Science Technology and Research (A*STAR) Singapore and College of Computing and Data Science NTU Singapore

Variance reduction techniques are designed to decrease the sampling variance, thereby accelerating convergence rates of first-order (FO) and zeroth-order (ZO) optimization methods. However, in composite optimization problems, ZO methods encounter an additional variance called the coordinate-wise variance, which stems from the random gradient estimation. To reduce this variance, prior works require estimating all partial derivatives, essentially approximating FO information. This approach demands O(d) function evaluations (d is the dimension size), which incurs substantial computational costs and is prohibitive in high-dimensional scenarios. This paper proposes the Zeroth-order Proximal Double Variance Reduction (ZPDVR) method, which utilizes the averaging trick to reduce both sampling and coordinate-wise variances. Compared to prior methods, ZPDVR relies solely on random gradient estimates, calls the stochastic zeroth-order oracle (SZO) in expectation O(1) times per iteration, and achieves the optimal O(d(n+κ) log(1/ε)) SZO query complexity in the strongly convex and smooth setting, where κ represents the condition number and ε is the desired accuracy. Empirical results validate ZPDVR's linear convergence and demonstrate its superior performance over other related methods.

关键词：

ReInform: Selecting paths with reinforcement learning for contextualized link prediction

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Speranskaya, Marina Methias, Sameh Roth, Benjamin Center for Information and Language Processing LMU Munich Germany Technical University of Munich Munich Germany Research Group Data Mining and Machine Learning University of Vienna Vienna Austria

We propose to use reinforcement learning to inform transformer-based contextualized link prediction models by providing paths that are most useful for predicting the correct answer. This is in contrast to previous approaches, that either used reinforcement learning (RL) to directly search for the answer, or based their prediction on limited or randomly selected context. Our experiments on WN18RR and FB15k-237 show that contextualized link prediction models consistently outperform RL-based answer search, and that additional improvements (of up to 13.5% MRR) can be gained by combining RL with a link prediction model. The PyTorch implementation of the RL agent is available at https://***/marina-sp/reinform. Copyright © 2022, The Authors. All rights reserved.

关键词： Reinforcement learning

Region-Aware Metric learning for Open World Semantic Segmentation via Meta-Channel Aggregation

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Dong, Hexin Chen, Zifan Yuan, Mingze Xie, Yutong Zhao, Jie Yu, Fei Dong, Bin Zhang, Li Center for Data Science Peking University Beijing China National Biomedical Imaging Center Peking University Beijing China Center for Machine Learning Research Peking University Beijing China Peking University Beijing China

As one of the most challenging and practical segmentation tasks, open-world semantic segmentation requires the model to segment the anomaly regions in the images and incrementally learn to segment out-of-distribution (OOD) objects, especially under a few-shot condition. The current state-of-the-art (SOTA) method, Deep Metric learning Network (DMLNet), relies on pixel-level metric learning, with which the identification of similar regions having different semantics is difficult. Therefore, we propose a method called region-aware metric learning (RAML), which first separates the regions of the images and generates region-aware features for further metric learning. RAML improves the integrity of the segmented anomaly regions. Moreover, we propose a novel meta-channel aggregation (MCA) module to further separate anomaly regions, forming high-quality sub-region candidates and thereby improving the model performance for OOD objects. To evaluate the proposed RAML, we have conducted extensive experiments and ablation studies on Lost And Found and Road Anomaly datasets for anomaly segmentation and the CityScapes dataset for incremental few-shot learning. The results show that the proposed RAML achieves SOTA performance in both stages of open world segmentation. Our code and appendix are available at https://***/czifan/RAML. Copyright © 2022, The Authors. All rights reserved.

关键词： Semantic Segmentation

Accelerating crystal structure search through active learning with neural networks for rapid relaxations

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Hessmann, Stefaan S.P. Schütt, Kristof T. Gebauer, Niklas W.A. Gastegger, Michael Oguchi, Tamio Yamashita, Tomoki Machine Learning Group Technische Universität Berlin Berlin10587 Germany Berlin Institute for the Foundations of Learning and Data Berlin10587 Germany BASLEARN TU Berlin BASF Joint Lab for Machine Learning Technische Universität Berlin Berlin10587 Germany Center for Spintronics Research Network Osaka University 1-3 Machikaneyama Toyonaka Osaka560-8531 Japan Department of Electrical Electronics and Information Engineering Nagaoka University of Technology 1603-1 Kamitomioka-machi Nagaoka Niigata940-2188 Japan

Global optimization of crystal compositions is a significant yet computationally intensive method to identify stable structures within chemical space. The specific physical properties linked to a three-dimensional atomic arrangement make this an essential task in the development of new materials. We present a method that efficiently uses active learning of neural network force fields for structure relaxation, minimizing the required number of steps in the process. This is achieved by neural network force fields equipped with uncertainty estimation, which iteratively guide a pool of randomly generated candidates towards their respective local minima. Using this approach, we are able to effectively identify the most promising candidates for further evaluation using density functional theory (DFT). Our method not only reliably reduces computational costs by up to two orders of magnitude across the benchmark systems Si16, Na8Cl8, Ga8As8 and Al4O6, but also excels in finding the most stable minimum for the unseen, more complex systems Si46 and Al16O24. Moreover, we demonstrate at the example of Si16 that our method can find multiple relevant local minima while only adding minor computational effort. Copyright © 2024, The Authors. All rights reserved.

关键词： Density functional theory

Design and Implementation of a data Governance Framework and Platform: A Case Study of a National research Organization of Thailand

学校读者我要写书评

暂无评论

Design and Implementation of a Data Governance Framework and...

International Joint Conference on Computer Science and Software Engineering (JCSSE)

作者： Sapa Chanyachatchawan Krich Nasingkun Patipat Tumsangthong Porntiwa Chata Marut Buranarach Monsak Socharoentum Leveraging Technology Solutions Section National Electronics and Computer Technology Center Bangkok Thailand Strategic Analytics Networks with Machine Learning and AI Research National Electronics and Computer Technology Center Bangkok Thailand Data Science and Analytics Research Group National Electronics and Computer Technology Center Bangkok Thailand Digital Government Development Agency Bangkok Thailand

In the current era of extensive data usage across industries, data collection, preservation, utilization, and organization has become more challenging and nuanced because it is necessary to consider critical concerns such as data security, privacy, and legal issues, apart from efficiency issues. As a result, Thai government initiated the idea and effort to implement data governance throughout the government agency. This paper showcases the implementation of data governance in a governmental research organization with highly diverse structured and unstructured data. The implementation follows international standards and the guidelines of the Digital Government Development Agency (DGA). The executives set up the working body, including the data Governance Council and data Stewards, responsible for setting up and deploying policies and regulations. Creating awareness and the necessary infrastructure are the main focuses in the first-year phase. The metadata was designed to extend DGA's version and match the organization's unique requirements. A data catalog platform was developed accordingly. We organized activities to boost employee awareness and participation, including advertising and data catalog platform training. By the end of the first year of implementation, every organization unit had registered at least one data record into the data catalog.

关键词：

On the Mathematics of RNA Velocity II: Algorithmic Aspects

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Li, Tiejun Wang, Yizhuo Yang, Guoguo Zhou, Peijie LMAM School of Mathematical Sciences Peking University Beijing100871 China Center for Machine Learning Research Peking University Beijing100871 China Center for Data Science Peking University Beijing100871 China Department of Mathematics University of California Irvine IrvineCA92697 United States

In a previous paper [CSIAM Trans. Appl. Math. 2 (2021), 1-55], the authors proposed a theoretical framework for the analysis of RNA velocity, which is a promising concept in scRNA-seq data analysis to reveal the cell state-transition dynamical processes underlying snapshot data. The current paper is devoted to the algorithmic study of some key components in RNA velocity workflow. Four important points are addressed in this paper: (1) We construct a rational time-scale fixation method which can determine the global gene-shared latent time for cells. (2) We present an uncertainty quantification strategy for the inferred parameters obtained through the EM algorithm. (3) We establish the optimal criterion for the choice of velocity kernel bandwidth with respect to the sample size in the downstream analysis and discuss its implications. (4) We propose a temporal distance estimation approach between two cell clusters along the cellular development path. Some illustrative numerical tests are also carried out to verify our analysis. These results are intended to provide tools and insights in further development of RNA velocity type methods in the future. © 2023, CC BY-NC-SA.

关键词： RNA