13,324 research outputs found
Multi-Stage Self-Supervised Learning for Graph Convolutional Networks on Graphs with Few Labeled Nodes
No full textGraph Convolutional Networks (GCNs) play a crucial role in graph learning tasks, however, learning graph embedding with few supervised signals is still a difficult problem. In this paper, we propose a novel training algorithm for Graph Convolutional Network, called Multi-Stage Self-Supervised (M3S) Training Algorithm, combined with self-supervised learning approach, focusing on improving the generalization performance of GCNs on graphs with few labeled nodes. Firstly, a Multi-Stage Training Framework is provided as the basis of M3S training method. Then we leverage DeepCluster technique, a popular form of self-supervised learning, and design corresponding aligning mechanism on the embedding space to refine the Multi-Stage Training Framework, resulting in M3S Training Algorithm. Finally, extensive experimental results verify the superior performance of our algorithm on graphs with few labeled nodes under different label rates compared with other state-of-the-art approaches
Dr. Lin Sun, CAU, March 2013
No full textThis video is a conversation with Dr. Lin Sun. Dr. Sun talks about an exhibit at the Woodruff Library titled "At The Boundary." Jordan Moore, AUC Woodruff Library, is the interviewer
An Analysis of <i>Judge Lin</i>
No full textBiography of Lin Wen Zhong Gong has another way to call, that is Judge Lin. The leading character is Lin Ze-Xu. This book is based on functionary experience of Lin Ze-Xu, with the captivating plots of court case, helping by highly skilled military attach\uc3\ua9s and chivalrous knights, and the history facts of Opium War. It makes Lin Ze-Xu\ue2s Confucian temperament and tragic mood more, also contrasts with author\ue2s sorrow and furiousness for the politics at the time. History, court case, martial arts\ue2\ua6\ue2\ua6etc. are essence of this book and it broadens the way of this writing style.
The topic of the thesis is \ue2An Analysis of Judge Lin\ue2. The following thesis will be divided into six different chapters. The introduction is Chapter one of the thesis, which is including researching motive and purpose, literature review of predecessors, researching version by existing information, raising questions, choosing research methods and arranging chapters. In chapter Two, I discuss the study of characters of Lin Ze-Xu, also makes a deep analysis of author\u27s purpose of writing him. In chapter Three, I analyze supporting actors and actress. Meanwhile, I illustrate author\u27s purpose of writing supporting actress because the author had different manner to describe supporting actress. Moving to the Chapter Four, I mainly focus on the plots of Judge Lin, and organize cases of Lin Ze-Xu and his subordinates to understand features of cases. In Chapter Five, I represent the causes of Opium War. China and England had difference of opinions of opium. Therefore, it is easier to comprehend what the author\u27s purpose is. In the last chapter I summarize the main points of the preceding chapters and confirm particularity of Judge Lin
L1-norm global geometric consistency for partial-duplicate image retrieval
No full textIn all feature point based partial-duplicate image retrieval systems, false matching is a common issue. To tackle the problem, geometric contexts are widely applied to filter the inconsistent matches. This paper presents a novel method called 1-norm global geometric consistency. We first form the squared distance matrices of all the matched feature points, which remain invariant under translation and rotation between partial-duplicated images. Then we find the scale difference by solving a one-variable 1-norm error minimization problem, where the large sparse errors correspond to the locations of inconsistent matches. By adopting the Golden Section Search method the minimization problem can be solved efficiently. Extensive experimental results show that our method reaches higher precisions than state-of-the-art geometric verification methods in detecting inconsistent matches. Its speed is also highly competitive even when compared to local geometric consistency based methods. ? 2014 IEEE.EI3033-303
Subspace clustering based tag sharing for inductive tag matrix refinement with complex errors
No full textAnnotating images with tags is useful for indexing and retrieving images. However, many available annotation data include missing or inaccurate annotations. In this paper, we propose an image annotation framework which sequentially performs tag completion and refinement. We utilize the subspace property of data via sparse subspace clustering for tag completion. Then we propose a novel matrix completion model for tag refinement, integrating visual correlation, semantic correlation and the novelly studied property of complex errors. The proposed method outperforms the state-of-the-art approaches on multiple benchmark datasets even when they contain certain levels of annotation noise. ? 2016 ACM.EI1013-101
Virtual adversarial training on graph convolutional networks in node classification
No full textThe effectiveness of Graph Convolutional Networks (GCNs) has been demonstrated in a wide range of graph-based machine learning tasks. However, the update of parameters in GCNs is only from labeled nodes, lacking the utilization of unlabeled data. In this paper, we apply Virtual Adversarial Training (VAT), an adversarial regularization method based on both labeled and unlabeled data, on the supervised loss of GCN to enhance its generalization performance. By imposing virtually adversarial smoothness on the posterior distribution in semi-supervised learning, VAT yields an improvement on the performance of GCNs. In addition, due to the difference of property in features, we perturb virtual adversarial perturbations on sparse and dense features, resulting in GCN Sparse VAT (GCNSVAT) and GCN Dense VAT (GCNDVAT) algorithms, respectively. Extensive experiments verify the effectiveness of our two methods across different training sizes. Our work paves the way towards better understanding the direction of improvement on GCNs in the future
Tensor LRR based subspace clustering
No full textSubspace clustering groups a set of samples (vectors) into clusters by approximating this set with a mixture of several linear subspaces, so that the samples in the same cluster are drawn from the same linear subspace. In majority of existing works on subspace clustering, samples are simply regarded as being independent and identically distributed, that is, arbitrarily ordering samples when necessary. However, this setting ignores sample correlations in their original spatial structure. To address this issue, we propose a tensor low-rank representation (TLRR) for subspace clustering by keeping available spatial information of data. TLRR seeks a lowest-rank representation over all the candidates while maintaining the inherent spatial structures among samples, and the affinity matrix used for spectral clustering is built from the combination of similarities along all data spatial directions. TLRR better captures the global structures of data and provides a robust subspace segmentation from corrupted data. Experimental results on both synthetic and real-world datasets show that TLRR outperforms several established state-of-the-art methods. ? 2014 IEEE.EI1877-188
Dual graph regularized latent low-rank representation for subspace clustering
No full textLow-rank representation (LRR) has received considerable attention in subspace segmentation due to its effectiveness in exploring low-dimensional subspace structures embedded in data. To preserve the intrinsic geometrical structure of data, a graph regularizer has been introduced into LRR framework for learning the locality and similarity information within data. However, it is often the case that not only the high-dimensional data reside on a non-linear low-dimensional manifold in the ambient space, but also their features lie on a manifold in feature space. In this paper, we propose a dual graph regularized LRR model (DGLRR) by enforcing preservation of geometric information in both the ambient space and the feature space. The proposed method aims for simultaneously considering the geometric structures of the data manifold and the feature manifold. Furthermore, we extend the DGLRR model to include non-negative constraint, leading to a parts-based representation of data. Experiments are conducted on several image data sets to demonstrate that the proposed method outperforms the state-of-the-art approaches in image clustering.Ming Yin, Junbin Gao, Zhouchen Lin, Qinfeng Shi, and Yi Gu
Appropriate Similarity Measures for Author Cocitation Analysis
No full textWe provide a number of new insights into the methodological discussion about author cocitation analysis. We first argue that the use of the Pearson correlation for measuring the similarity between authors’ cocitation profiles is not very satisfactory. We then discuss what kind of similarity measures may be used as an alternative to the Pearson correlation. We consider three similarity measures in particular. One is the well-known cosine. The other two similarity measures have not been used before in the bibliometric literature. Finally, we show by means of an example that our findings have a high practical relevance.information science;Pearson correlation;cosine;similarity measure;author cocitation analysis
Lowest electronic states of neutral and ionic LiN
No full textWe have investigated the potential energy curves (PECs) of the LiN heteronuclear diatomic molecule, including its ionic species LiN+ and LiN−, using explicitly correlated multi-reference configuration interaction (MRCI-F12) calculations in conjunction with the correlation consistent quintuple- basis set. The effect of core–valence correlation, scalar relativistic effects, and the size of the basis sets has been investigated. A comprehensive set of spectroscopic constants determined based on the above-mentioned calculations are also reported for the lowest electronic states and all systems, including dissociation energies, harmonic and anharmonic vibrational frequencies, and rotational constants. Additional parameters, such as the dipole moments, equilibrium spin-orbit constants, excitation energies, and rovibrational energy levels, are also documented. We found that the three triplet states of LiN, namely, X 3∑−, A 3Π, and 2 3∑−, exhibit substantial potential wells in the PEC diagrams, while the quintet states are repulsive in nature. The ground state of the anion also shows a deep potential well in the vicinity of its equilibrium geometry. In contrast, the ground and excited states of the cation are very loosely bound. Charge transfer properties of each of these states are also analyzed to obtain an in-depth understanding of the interatomic interactions. We found that the core–valence correlation has a substantial effect on the calculated spectroscopic constants.Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.Atmospheric Remote Sensin
- …
