检索结果-内蒙古大学图书馆

arXiv 2022年

作者： Ye, Zihan Shindo, Hikaru Dhami, Devendra Singh Kersting, Kristian AI and Machine Learning Group Dept. of Computer Science TU Darmstadt Germany Centre for Cognitive Science TU Darmstadt Germany Germany Germany Eindhoven University of Technology Netherlands

Deep neural learning uses an increasing amount of computation and data to solve very specific problems. By stark contrast, human minds solve a wide range of problems using a fixed amount of computation and limited experience. One ability that seems crucial to this kind of general intelligence is meta-reasoning, i.e., our ability to reason about reasoning. To make deep learning do more from less, we propose the first neural meta-symbolic system (NEMESYS) for reasoning and learning: meta programming using differentiable forward-chaining reasoning in first-order logic. Differentiable meta programming naturally allows NEMESYS to reason and learn several tasks efficiently. This is different from performing object-level deep reasoning and learning, which refers in some way to entities external to the system. In contrast, NEMESYS enables self-introspection, lifting from object- to meta-level reasoning and vice versa. In our extensive experiments, we demonstrate that NEMESYS can solve different kinds of tasks by adapting the meta-level programs without modifying the internal reasoning system. Moreover, we show that NEMESYS can learn meta-level programs given examples. This is difficult, if not impossible, for standard differentiable logic programming. Copyright © 2022, The Authors. All rights reserved.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Multi-View Brain Network Analysis with Cross-View Missing Network Generation

Multi-View Brain Network Analysis with Cross-View Missing Ne...

引用

IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

作者： Gongxu Luo Chenyang Li Hejie Cui Lichao Sun Lifang He Carl Yang Department of Machine Learning MBZUAI Abu Dhabi United Arab Emirates Dept. of Electrical and Computer Engineering University of Waterloo Waterloo ON Canada Department of Computer Science Emory University Atlanta GA United States Dept. of Computer Science and Engineering Lehigh University Bethlehem PA United States

ISBN: (数字)9781665468190

ISBN: (纸本)9781665468206

Parkinson’s Disease (PD), one of the most common neurological disorders, has long been a challenge in public health clinical diagnosis as well as scientific understanding. Recently, there has been an upsurge of interest in brain network analysis which benefits the understanding of brain functions and early detection of neurological disorders extensively. Multi-view brain networks with different connectivity patterns among regions of interests (ROIs) can be constructed to reflect different and complementary perspectives of the brain connectivity profile. However, the extraction of such multi-view brain networks relies on the availability of multiple neuroimaging modalities and heavy data preprocessing, which often leads to severe missing data in either view. The cross-view missing issue hinders the pragmaticality of multi-view representation learning and downstream analysis. In this work, we formulate the novel problem of cross-view brain network generation and propose CroGen, a graph generative model that can generate the missing view when only one view is given. Specifically, GroGen leverages the potential correlation between diverse views of brain networks of the same individuals. Moreover, we design a pre-train schema to make full use of the labeled individuals with only single views of brain networks. Extensive experiments on real-life Parkinson’s Progression Markers Initiative (PPMI) cohort demonstrate the supreme effectiveness of CroGen over baselines on both cross-view generation tasks and downstream PD classification.

关键词： Neurological diseases Representation learning Neuroimaging Knowledge engineering Data preprocessing Network analyzers Brain modeling

来源：评论

学校读者我要写书评

暂无评论

Graph Filtration Kernels

arXiv

引用

arXiv 2021年

作者： Schulz, Till Hendrik Welke, Pascal Wrobel, Stefan Dept. of Computer Science Universität Bonn Germany Center for Machine Learning Sankt Augustin Germany

The majority of popular graph kernels is based on the concept of Haussler’s R-convolution kernel and defines graph similarities in terms of mutual substructures. In this work, we enrich these similarity measures by considering graph filtrations: Using meaningful orders on the set of edges, which allow to construct a sequence of nested graphs, we can consider a graph at multiple granularities. For one thing, this provides access to features on different levels of resolution. Furthermore, rather than to simply compare frequencies of features in graphs, it allows for their comparison in terms of when and for how long they exist in the sequences. In this work, we propose a family of graph kernels that incorporate these existence intervals of features. While our approach can be applied to arbitrary graph features, we particularly highlight Weisfeiler-Lehman vertex labels, leading to efficient kernels. We show that using Weisfeiler-Lehman labels over certain filtrations strictly increases the expressive power over the ordinary Weisfeiler-Lehman procedure in terms of deciding graph isomorphism. In fact, this result directly yields more powerful graph kernels based on such features and has implications to graph neural networks due to their close relationship to the Weisfeiler-Lehman method. We empirically validate the expressive power of our graph kernels and show significant improvements over state-of-the-art graph kernels in terms of predictive performance on various real-world benchmark datasets. Copyright © 2021, The Authors. 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

A Fast Heuristic for Computing Geodesic Cores in Large Networks

arXiv

引用

arXiv 2022年

作者： Seiffarth, Florian Horváth, Tamás Wrobel, Stefan Dept. of Computer Science University of Bonn Bonn Germany Fraunhofer IAIS Schloß Birlinghoven Sankt Augustin Germany Fraunhofer Center for Machine Learning Schloß Birlinghoven Sankt Augustin Germany

Motivated by the increasing interest in applications of graph geodesic convexity in machine learning and data mining, we present a heuristic for computing the geodesic convex hull of node sets in networks. It generates a set of almost maximal outerplanar spanning subgraphs for the input graph, computes the geodesic closure in each of these graphs, and regards a node as an element of the convex hull if it belongs to the closed sets for at least a user specified number of outerplanar graphs. Our heuristic algorithm runs in time linear in the number of edges of the input graph, i.e., it is faster with one order of magnitude than the standard algorithm computing the closure exactly. Its performance is evaluated empirically by approximating convexity based core-periphery decomposition of networks. Our experimental results with large real-world networks show that for most networks, the proposed heuristic was able to produce close approximations significantly faster than the standard algorithm computing the exact convex hulls. For example, while our algorithm calculated an approximate core-periphery decomposition in 5 hours or less for networks with more than 20 million edges, the standard algorithm did not terminate within 50 days. Copyright © 2022, The Authors. All rights reserved.

关键词： Heuristic algorithms

来源：评论

学校读者我要写书评

暂无评论

On the Non-Computability of Convex Optimization Problems

On the Non-Computability of Convex Optimization Problems

引用

IEEE International Symposium on Information Theory

作者： Holger Boche Andrea Grigorescu Rafael F. Schaefer H. Vincent Poor Chair of Theoretical Information Technology Technical University of Munich BMBF Research Hub 6G-life Chair of Information Theory and Machine Learning Technische Universität Dresden BMBF Research Hub 6G-life CeTI - Excellence Cluster Dept. of Electrical and Computer Engineering Princeton University

ISBN: (数字)9798350382846

ISBN: (纸本)9798350382853

This paper explores the computability of the optimal point in convex problems with inequality constraints. It is shown that feasible sets, defined by computable convex functions, can yield non-computable optimal points for strictly convex and computable objective functions. Additionally, the optimal point of the Lagrangian dual problem associated with such convex constraints is also proven to be non-computable. Despite converging sequences of computable numbers towards the Lagrangian's optimal point, algorithmic control of the approximation error is shown to be impossible.

关键词： Approximation algorithms Linear programming Approximation error Convex functions Optimization Information theory

来源：评论

学校读者我要写书评

暂无评论

Image Copy Move Forgery Detection Using Multi-Plane Convolutional Neural Network

Image Copy Move Forgery Detection Using Multi-Plane Convolut...

引用

Distributed Computing and Optimization Techniques (ICDCOT), International Conference on

作者： S Vagdevi Deepthi V S Shweta Dhareshwar Mohammed Al-Farouni Hanumanthakari Kalyan Rao Artificial Intelligence and Machine Learning City Engineering College Bangalore Karnataka India Computer Science and Engineering (Cyber Security) Dayananda Sagar college of engineering Bangalore Karnataka India Department of MCA Nitte Meenakshi Institute of Technology Bengaluru India The Islamic university Najaf Iraq Dept. of CSE GRIET Hyderabad Telangana India

In digital images, the digital device usage and availability of open source for image editing applications are easily manipulating the digital images. The Copy Move Forgery (CMF) is an extensive technique for duplicate or hide particular image portion without leaving visual signs. Therefore, it is complex to detect the CMF and the forensic experts are rely on efficient technique for CMF detection. The Multi Plane Convolutional Neural Network (MP-CNN) is proposed in this research for detecting CMF in images. The MICC-F220, MICC-F2000 and CASIA 2.0 datasets are used which is publicly available Kaggle dataset. This dataset is preprocessed by image denoising which reduces the noise from input image. Then, the pre-processed images are provided to MP-CNN for detecting the CMF. The metrics such as precision, accuracy, f1score and recall are applied for estimating the MP-CNN performance. The MP-CNN attains accuracy of 99.45%, 99.26% and 98.61% for MICC-F220, MICC-F2000 and CASIA 2.0 datasets when compared to existing techniques like CNN, Masked Region CNN (M-RCNN) and Contrast Limited Adaptive Histogram Equalization with CNN (CLAHE+CNN).

关键词：

来源：评论

学校读者我要写书评

暂无评论

DEEP ENSEMBLE FOR ROTORCRAFT ATTITUDE PREDICTION

arXiv

引用

arXiv 2023年

作者： Khan, Hikmat Bouaynaya, Nidhal C. Rasool, Ghulam Travis, Tyler Thompson, Lacey Johnson, Charles C. Dept. of Electrical and Computer Engineering Rowan University GlassboroNJ United States Dept. of Machine Learning Moffitt Cancer Center TampaFL United States Federal Aviation Administration Atlantic CityNJ United States Research Engineer Flight Test Engineer and Program Lead Federal Aviation Administration Egg Harbor Township United Kingdom

Historically, the rotorcraft community has experienced a higher fatal accident rate than other aviation segments, including commercial and general aviation. To date, traditional methods applied to reduce incident rates have not proven hugely successful for the rotorcraft community. Recent advancements in artificial intelligence (AI) and the application of these technologies in different areas of our lives are both intriguing and encouraging. When developed appropriately for the aviation domain, AI techniques may provide an opportunity to help design systems that can address rotorcraft safety challenges. Our recent work demonstrated that AI algorithms could use video data from onboard cameras and correctly identify different flight parameters from cockpit gauges, e.g., indicated airspeed. These AI-based techniques provide a potentially cost-effective solution, especially for small helicopter operators, to record the flight state information and perform post-flight analyses. We also showed that carefully designed and trained AI systems can accurately predict rotorcraft attitude (i.e., pitch and yaw) from outside scenes (images or video data). Ordinary off-the-shelf video cameras were installed inside the rotorcraft cockpit to record the outside scene, including the horizon. The AI algorithm was able to correctly identify rotorcraft attitude at an accuracy in the range of 80%. In this work, we combined five different onboard camera viewpoints to improve attitude prediction accuracy to 94%. Our current approach, which is referred to as ensembled prediction, significantly increased the reliability in the predicted attitude (i.e., pitch and yaw). For example, in some camera views, the horizon may be obstructed or not visible. The proposed ensemble method can combine visual details recorded from other cameras and predict the attitude with high reliability. In our setup, the five onboard camera views included pilot windshield, co-pilot windshield, pilot Electronic Flight In

关键词： Instrument displays

来源：评论

学校读者我要写书评

暂无评论

From Lenses to Living Rooms: A Policy Brief on Eye Tracking in XR Before the Impending Boom

From Lenses to Living Rooms: A Policy Brief on Eye Tracking ...

引用

IEEE International Conference on Artificial Intelligence and Virtual Reality (AIVR)

作者： Sean Anthony Byrne Nora Castner Efe Bozkir Diederick C. Niehorster Enkelejda Kasneci MoMiLab IMT School for Advanced Studies Lucca Lucca Italy Carl Zeiss Vision International GmbH Tübingen Germany Human-Centered Technologies for Learning Technical University of Munich Munich Germany Human-Computer Interaction University of Tübingen Tübingen Germany Dept. of Psychology Lund University Humanities Lab & Lund University Lund Sweden

As tech giants such as Apple and Meta invest heavily in Virtual and Augmented Reality (VR/AR) technologies, often collectively termed Extended Reality (XR) devices, a significant societal concern emerges: The use of eye-tracking technology within these devices. Gaze data holds immense value, revealing insights into user attention, health, and cognitive states. This raises substantial concerns over privacy and fairness, with potential risks of targeted ads, unauthorized surveillance, and data re-purposing. As the impact of eye tracking is due to transition from the lab to the broader public, this paper underscores these pivotal issues in a digestible manner to the general audience. To this end, we first outline the eye-tracking data collection process and its potential for user insights. Second, we introduce perspectives from the domain of privacy, emphasizing its significance as a pivotal measure to guard against the improper use of eye-tracking data. Third, we provide a set of guidelines created by researchers actively working within this space. These recommendations are designed to guide policymakers and the general public toward establishing informed, equitable, and privacy-centric standards surrounding these devices.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Improving language generation from feature-rich tree-structured data with relational graph convolutional encoders 2

Improving language generation from feature-rich tree-structu...

引用

2nd Workshop on Multilingual Surface Realisation, MSR@EMNLP-IJCNLP 2019

作者： Hong, Xudong Chang, Ernie Demberg, Vera Dept. of Computer Vision and Machine Learning MPI Informatics Germany Dept. of Language Science and Technology Saarland University Germany

ISBN: (纸本)9781950737765

The Multilingual Surface Realization Shared Task 2019 focuses on generating sentences from lemmatized sets of universal dependency parses with rich features. This paper describes the system design and the results of our participation in the deep track. The core innovation in our approach is to use a graph convolutional network to encode the dependency trees given as input. Upon adding morphological features, our system achieves the second rank in the deep track without using data augmentation techniques or additional components (such as a re-ranker). © 2019 Association for Computational Linguistics.

关键词： Forestry

来源：评论

学校读者我要写书评

暂无评论

Characterization of the Complexity of Computing the Capacity of Colored Noise Gaussian Channels

Characterization of the Complexity of Computing the Capacity...

引用

IEEE International Conference on Communications (ICC)

作者： Holger Boche Andrea Grigorescu Rafael F. Schaefer H. Vincent Poor Chair of Theoretical Information Technology Technical University of Munich BMBF Research Hub 6G-life Munich Quantum Valley (MQV) Chair of Information Theory and Machine Learning Technische Universität Dresden BMBF Research Hub 6G-life Cluster of Excellence CeTI Dept. of Electrical and Computer Engineering Princeton University

ISBN: (数字)9781728190549

ISBN: (纸本)9781728190556

This paper investigates the computational complexity involved in determining the capacity of the band-limited additive colored Gaussian noise (ACGN) channel and its capacity-achieving input power spectral density (p.s.d.). A band-limited polynomial time computable continuous and strictly positive noise p.s.d. is constructed for the ACGN channel such that the computation of its corresponding capacity is $\# \mathrm{P}_{1}$ -complete. This means that it is even more complex than problems that are $\text{NP}_{1}$ -complete. Additionally, it is shown that computing the capacity-achieving input p.s.d. is also $\# \mathrm{P}_{1}$ -complete. Furthermore, under the widely accepted assumption that $\text{FP}_{1}\neq\# \mathrm{P}_{1}$ , there are two significant implications for the ACGN channel. First, there exists a polynomial time computable noise p.s.d. for which computing its capacity is not polynomial-time feasible, meaning the number of computational steps on a Turing machine grows faster than any polynomial. Second, there is a polynomial time computable noise p.s.d. where determining its capacity-achieving input p.s.d. is also not achievable in polynomial time.

关键词： Additives Turing machines Gaussian noise Polynomials Computational complexity Colored noise Gaussian channels

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：