July 28, 2019

3242 words 16 mins read

Paper Group ANR 343

Zero-shot Learning via Shared-Reconstruction-Graph Pursuit. Grade Prediction with Temporal Course-wise Influence. 3D Randomized Connection Network with Graph-based Label Inference. Adversarial Networks for Prostate Cancer Detection. Transformation Models in High-Dimensions. Phase Transitions of Spectral Initialization for High-Dimensional Nonconvex …

Zero-shot Learning via Shared-Reconstruction-Graph Pursuit


Title	Zero-shot Learning via Shared-Reconstruction-Graph Pursuit
Authors	Bo Zhao, Xinwei Sun, Yuan Yao, Yizhou Wang
Abstract	Zero-shot learning (ZSL) aims to recognize objects from novel unseen classes without any training data. Recently, structure-transfer based methods are proposed to implement ZSL by transferring structural knowledge from the semantic embedding space to image feature space to classify testing images. However, we observe that such a knowledge transfer framework may suffer from the problem of the geometric inconsistency between the data in the training and testing spaces. We call this problem as the space shift problem. In this paper, we propose a novel graph based method to alleviate this space shift problem. Specifically, a Shared Reconstruction Graph (SRG) is pursued to capture the common structure of data in the two spaces. With the learned SRG, each unseen class prototype (cluster center) in the image feature space can be synthesized by the linear combination of other class prototypes, so that testing instances can be classified based on the distance to these synthesized prototypes. The SRG bridges the image feature space and semantic embedding space. By applying spectral clustering on the learned SRG, many meaningful clusters can be discovered, which interprets ZSL performance on the datasets. Our method can be easily extended to the generalized zero-shot learning setting. Experiments on three popular datasets show that our method outperforms other methods on all datasets. Even with a small number of training samples, our method can achieve the state-of-the-art performance.
Tasks	Transfer Learning, Zero-Shot Learning
Published	2017-11-20
URL	http://arxiv.org/abs/1711.07302v1
PDF	http://arxiv.org/pdf/1711.07302v1.pdf
PWC	https://paperswithcode.com/paper/zero-shot-learning-via-shared-reconstruction
Repo
Framework

Grade Prediction with Temporal Course-wise Influence


Title	Grade Prediction with Temporal Course-wise Influence
Authors	Zhiyun Ren, Xia Ning, Huzefa Rangwala
Abstract	There is a critical need to develop new educational technology applications that analyze the data collected by universities to ensure that students graduate in a timely fashion (4 to 6 years); and they are well prepared for jobs in their respective fields of study. In this paper, we present a novel approach for analyzing historical educational records from a large, public university to perform next-term grade prediction; i.e., to estimate the grades that a student will get in a course that he/she will enroll in the next term. Accurate next-term grade prediction holds the promise for better student degree planning, personalized advising and automated interventions to ensure that students stay on track in their chosen degree program and graduate on time. We present a factorization-based approach called Matrix Factorization with Temporal Course-wise Influence that incorporates course-wise influence effects and temporal effects for grade prediction. In this model, students and courses are represented in a latent “knowledge” space. The grade of a student on a course is modeled as the similarity of their latent representation in the “knowledge” space. Course-wise influence is considered as an additional factor in the grade prediction. Our experimental results show that the proposed method outperforms several baseline approaches and infer meaningful patterns between pairs of courses within academic programs.
Tasks
Published	2017-09-15
URL	http://arxiv.org/abs/1709.05433v1
PDF	http://arxiv.org/pdf/1709.05433v1.pdf
PWC	https://paperswithcode.com/paper/grade-prediction-with-temporal-course-wise
Repo
Framework

3D Randomized Connection Network with Graph-based Label Inference


Title	3D Randomized Connection Network with Graph-based Label Inference
Authors	Siqi Bao, Pei Wang, Tony C. W. Mok, Albert C. S. Chung
Abstract	In this paper, a novel 3D deep learning network is proposed for brain MR image segmentation with randomized connection, which can decrease the dependency between layers and increase the network capacity. The convolutional LSTM and 3D convolution are employed as network units to capture the long-term and short-term 3D properties respectively. To assemble these two kinds of spatial-temporal information and refine the deep learning outcomes, we further introduce an efficient graph-based node selection and label inference method. Experiments have been carried out on two publicly available databases and results demonstrate that the proposed method can obtain competitive performances as compared with other state-of-the-art methods.
Tasks	Semantic Segmentation
Published	2017-11-11
URL	http://arxiv.org/abs/1711.04170v1
PDF	http://arxiv.org/pdf/1711.04170v1.pdf
PWC	https://paperswithcode.com/paper/3d-randomized-connection-network-with-graph
Repo
Framework

Adversarial Networks for Prostate Cancer Detection


Title	Adversarial Networks for Prostate Cancer Detection
Authors	Simon Kohl, David Bonekamp, Heinz-Peter Schlemmer, Kaneschka Yaqubi, Markus Hohenfellner, Boris Hadaschik, Jan-Philipp Radtke, Klaus Maier-Hein
Abstract	The large number of trainable parameters of deep neural networks renders them inherently data hungry. This characteristic heavily challenges the medical imaging community and to make things even worse, many imaging modalities are ambiguous in nature leading to rater-dependant annotations that current loss formulations fail to capture. We propose employing adversarial training for segmentation networks in order to alleviate aforementioned problems. We learn to segment aggressive prostate cancer utilizing challenging MRI images of 152 patients and show that the proposed scheme is superior over the de facto standard in terms of the detection sensitivity and the dice-score for aggressive prostate cancer. The achieved relative gains are shown to be particularly pronounced in the small dataset limit.
Tasks
Published	2017-11-28
URL	http://arxiv.org/abs/1711.10400v1
PDF	http://arxiv.org/pdf/1711.10400v1.pdf
PWC	https://paperswithcode.com/paper/adversarial-networks-for-prostate-cancer
Repo
Framework

Transformation Models in High-Dimensions


Title	Transformation Models in High-Dimensions
Authors	Sven Klaassen, Jannis Kueck, Martin Spindler
Abstract	Transformation models are a very important tool for applied statisticians and econometricians. In many applications, the dependent variable is transformed so that homogeneity or normal distribution of the error holds. In this paper, we analyze transformation models in a high-dimensional setting, where the set of potential covariates is large. We propose an estimator for the transformation parameter and we show that it is asymptotically normally distributed using an orthogonalized moment condition where the nuisance functions depend on the target parameter. In a simulation study, we show that the proposed estimator works well in small samples. A common practice in labor economics is to transform wage with the log-function. In this study, we test if this transformation holds in CPS data from the United States.
Tasks
Published	2017-12-20
URL	http://arxiv.org/abs/1712.07364v1
PDF	http://arxiv.org/pdf/1712.07364v1.pdf
PWC	https://paperswithcode.com/paper/transformation-models-in-high-dimensions
Repo
Framework

Phase Transitions of Spectral Initialization for High-Dimensional Nonconvex Estimation


Title	Phase Transitions of Spectral Initialization for High-Dimensional Nonconvex Estimation
Authors	Yue M. Lu, Gen Li
Abstract	We study a spectral initialization method that serves a key role in recent work on estimating signals in nonconvex settings. Previous analysis of this method focuses on the phase retrieval problem and provides only performance bounds. In this paper, we consider arbitrary generalized linear sensing models and present a precise asymptotic characterization of the performance of the method in the high-dimensional limit. Our analysis also reveals a phase transition phenomenon that depends on the ratio between the number of samples and the signal dimension. When the ratio is below a minimum threshold, the estimates given by the spectral method are no better than random guesses drawn from a uniform distribution on the hypersphere, thus carrying no information; above a maximum threshold, the estimates become increasingly aligned with the target signal. The computational complexity of the method, as measured by the spectral gap, is also markedly different in the two phases. Worked examples and numerical results are provided to illustrate and verify the analytical predictions. In particular, simulations show that our asymptotic formulas provide accurate predictions for the actual performance of the spectral method even at moderate signal dimensions.
Tasks
Published	2017-02-21
URL	https://arxiv.org/abs/1702.06435v3
PDF	https://arxiv.org/pdf/1702.06435v3.pdf
PWC	https://paperswithcode.com/paper/phase-transitions-of-spectral-initialization
Repo
Framework

Towards Bursting Filter Bubble via Contextual Risks and Uncertainties


Title	Towards Bursting Filter Bubble via Contextual Risks and Uncertainties
Authors	Rikiya Takahashi, Shunan Zhang
Abstract	A rising topic in computational journalism is how to enhance the diversity in news served to subscribers to foster exploration behavior in news reading. Despite the success of preference learning in personalized news recommendation, their over-exploitation causes filter bubble that isolates readers from opposing viewpoints and hurts long-term user experiences with lack of serendipity. Since news providers can recommend neither opposite nor diversified opinions if unpopularity of these articles is surely predicted, they can only bet on the articles whose forecasts of click-through rate involve high variability (risks) or high estimation errors (uncertainties). We propose a novel Bayesian model of uncertainty-aware scoring and ranking for news articles. The Bayesian binary classifier models probability of success (defined as a news click) as a Beta-distributed random variable conditional on a vector of the context (user features, article features, and other contextual features). The posterior of the contextual coefficients can be computed efficiently using a low-rank version of Laplace’s method via thin Singular Value Decomposition. Efficiencies in personalized targeting of exceptional articles, which are chosen by each subscriber in test period, are evaluated on real-world news datasets. The proposed estimator slightly outperformed existing training and scoring algorithms, in terms of efficiency in identifying successful outliers.
Tasks
Published	2017-06-30
URL	http://arxiv.org/abs/1706.09985v1
PDF	http://arxiv.org/pdf/1706.09985v1.pdf
PWC	https://paperswithcode.com/paper/towards-bursting-filter-bubble-via-contextual
Repo
Framework


Title	Low-rank and Sparse NMF for Joint Endmembers’ Number Estimation and Blind Unmixing of Hyperspectral Images
Authors	Paris V. Giampouras, Athanasios A. Rontogiannis, Konstantinos D. Koutroumbas
Abstract	Estimation of the number of endmembers existing in a scene constitutes a critical task in the hyperspectral unmixing process. The accuracy of this estimate plays a crucial role in subsequent unsupervised unmixing steps i.e., the derivation of the spectral signatures of the endmembers (endmembers’ extraction) and the estimation of the abundance fractions of the pixels. A common practice amply followed in literature is to treat endmembers’ number estimation and unmixing, independently as two separate tasks, providing the outcome of the former as input to the latter. In this paper, we go beyond this computationally demanding strategy. More precisely, we set forth a multiple constrained optimization framework, which encapsulates endmembers’ number estimation and unsupervised unmixing in a single task. This is attained by suitably formulating the problem via a low-rank and sparse nonnegative matrix factorization rationale, where low-rankness is promoted with the use of a sophisticated $\ell_2/\ell_1$ norm penalty term. An alternating proximal algorithm is then proposed for minimizing the emerging cost function. The results obtained by simulated and real data experiments verify the effectiveness of the proposed approach.
Tasks	Hyperspectral Unmixing
Published	2017-03-16
URL	http://arxiv.org/abs/1703.05785v1
PDF	http://arxiv.org/pdf/1703.05785v1.pdf
PWC	https://paperswithcode.com/paper/low-rank-and-sparse-nmf-for-joint-endmembers
Repo
Framework

Inertia-Constrained Pixel-by-Pixel Nonnegative Matrix Factorisation: a Hyperspectral Unmixing Method Dealing with Intra-class Variability


Title	Inertia-Constrained Pixel-by-Pixel Nonnegative Matrix Factorisation: a Hyperspectral Unmixing Method Dealing with Intra-class Variability
Authors	Charlotte Revel, Yannick Deville, Véronique Achard, Xavier Briottet
Abstract	Blind source separation is a common processing tool to analyse the constitution of pixels of hyperspectral images. Such methods usually suppose that pure pixel spectra (endmembers) are the same in all the image for each class of materials. In the framework of remote sensing, such an assumption is no more valid in the presence of intra-class variabilities due to illumination conditions, weathering, slight variations of the pure materials, etc… In this paper, we first describe the results of investigations highlighting intra-class variability measured in real images. Considering these results, a new formulation of the linear mixing model is presented leading to two new methods. Unconstrained Pixel-by-pixel NMF (UP-NMF) is a new blind source separation method based on the assumption of a linear mixing model, which can deal with intra-class variability. To overcome UP-NMF limitations an extended method is proposed, named Inertia-constrained Pixel-by-pixel NMF (IP-NMF). For each sensed spectrum, these extended versions of NMF extract a corresponding set of source spectra. A constraint is set to limit the spreading of each source’s estimates in IP-NMF. The methods are tested on a semi-synthetic data set built with spectra extracted from a real hyperspectral image and then numerically mixed. We thus demonstrate the interest of our methods for realistic source variabilities. Finally, IP-NMF is tested on a real data set and it is shown to yield better performance than state of the art methods.
Tasks	Hyperspectral Unmixing
Published	2017-02-24
URL	http://arxiv.org/abs/1702.07630v1
PDF	http://arxiv.org/pdf/1702.07630v1.pdf
PWC	https://paperswithcode.com/paper/inertia-constrained-pixel-by-pixel
Repo
Framework

Semi-Supervised Endmember Identification In Nonlinear Spectral Mixtures Via Semantic Representation


Title	Semi-Supervised Endmember Identification In Nonlinear Spectral Mixtures Via Semantic Representation
Authors	Yuki Itoh, Siwei Feng, Marco F. Duarte, Mario Parente
Abstract	This paper proposes a new hyperspectral unmixing method for nonlinearly mixed hyperspectral data using a semantic representation in a semi-supervised fashion, assuming the availability of a spectral reference library. Existing semi-supervised unmixing algorithms select members from an endmember library that are present at each of the pixels; most such methods assume a linear mixing model. However, those methods will fail in the presence of nonlinear mixing among the observed spectra. To address this issue, we develop an endmember selection method using a recently proposed semantic spectral representation obtained via non-homogeneous hidden Markov chain (NHMC) model for a wavelet transform of the spectra. The semantic representation can encode spectrally discriminative features for any observed spectrum and, therefore, our proposed method can perform endmember selection without any assumption on the mixing model. Experimental results show that in the presence of sufficiently nonlinear mixing our proposed method outperforms dictionary-based sparse unmixing approaches based on linear models.
Tasks	Hyperspectral Unmixing
Published	2017-01-03
URL	http://arxiv.org/abs/1701.00804v1
PDF	http://arxiv.org/pdf/1701.00804v1.pdf
PWC	https://paperswithcode.com/paper/semi-supervised-endmember-identification-in
Repo
Framework

Semantic Softmax Loss for Zero-Shot Learning


Title	Semantic Softmax Loss for Zero-Shot Learning
Authors	Zhong Ji, Yunxin Sun, Yulong Yu, Jichang Guo, Yanwei Pang
Abstract	A typical pipeline for Zero-Shot Learning (ZSL) is to integrate the visual features and the class semantic descriptors into a multimodal framework with a linear or bilinear model. However, the visual features and the class semantic descriptors locate in different structural spaces, a linear or bilinear model can not capture the semantic interactions between different modalities well. In this letter, we propose a nonlinear approach to impose ZSL as a multi-class classification problem via a Semantic Softmax Loss by embedding the class semantic descriptors into the softmax layer of multi-class classification network. To narrow the structural differences between the visual features and semantic descriptors, we further use an L2 normalization constraint to the differences between the visual features and visual prototypes reconstructed with the semantic descriptors. The results on three benchmark datasets, i.e., AwA, CUB and SUN demonstrate the proposed approach can boost the performances steadily and achieve the state-of-the-art performance for both zero-shot classification and zero-shot retrieval.
Tasks	Zero-Shot Learning
Published	2017-05-22
URL	http://arxiv.org/abs/1705.07692v1
PDF	http://arxiv.org/pdf/1705.07692v1.pdf
PWC	https://paperswithcode.com/paper/semantic-softmax-loss-for-zero-shot-learning
Repo
Framework

Individualized Risk Prognosis for Critical Care Patients: A Multi-task Gaussian Process Model


Title	Individualized Risk Prognosis for Critical Care Patients: A Multi-task Gaussian Process Model
Authors	Ahmed M. Alaa, Jinsung Yoon, Scott Hu, Mihaela van der Schaar
Abstract	We report the development and validation of a data-driven real-time risk score that provides timely assessments for the clinical acuity of ward patients based on their temporal lab tests and vital signs, which allows for timely intensive care unit (ICU) admissions. Unlike the existing risk scoring technologies, the proposed score is individualized; it uses the electronic health record (EHR) data to cluster the patients based on their static covariates into subcohorts of similar patients, and then learns a separate temporal, non-stationary multi-task Gaussian Process (GP) model that captures the physiology of every subcohort. Experiments conducted on data from a heterogeneous cohort of 6,094 patients admitted to the Ronald Reagan UCLA medical center show that our risk score significantly outperforms the state-of-the-art risk scoring technologies, such as the Rothman index and MEWS, in terms of timeliness, true positive rate (TPR), and positive predictive value (PPV). In particular, the proposed score increases the AUC with 20% and 38% as compared to Rothman index and MEWS respectively, and can predict ICU admissions 8 hours before clinicians at a PPV of 35% and a TPR of 50%. Moreover, we show that the proposed risk score allows for better decisions on when to discharge clinically stable patients from the ward, thereby improving the efficiency of hospital resource utilization.
Tasks
Published	2017-05-22
URL	http://arxiv.org/abs/1705.07674v1
PDF	http://arxiv.org/pdf/1705.07674v1.pdf
PWC	https://paperswithcode.com/paper/individualized-risk-prognosis-for-critical
Repo
Framework

Why & When Deep Learning Works: Looking Inside Deep Learnings


Title	Why & When Deep Learning Works: Looking Inside Deep Learnings
Authors	Ronny Ronen
Abstract	The Intel Collaborative Research Institute for Computational Intelligence (ICRI-CI) has been heavily supporting Machine Learning and Deep Learning research from its foundation in 2012. We have asked six leading ICRI-CI Deep Learning researchers to address the challenge of “Why & When Deep Learning works”, with the goal of looking inside Deep Learning, providing insights on how deep networks function, and uncovering key observations on their expressiveness, limitations, and potential. The output of this challenge resulted in five papers that address different facets of deep learning. These different facets include a high-level understating of why and when deep networks work (and do not work), the impact of geometry on the expressiveness of deep networks, and making deep networks interpretable.
Tasks
Published	2017-05-10
URL	http://arxiv.org/abs/1705.03921v1
PDF	http://arxiv.org/pdf/1705.03921v1.pdf
PWC	https://paperswithcode.com/paper/why-when-deep-learning-works-looking-inside
Repo
Framework

Gradient Coding from Cyclic MDS Codes and Expander Graphs


Title	Gradient Coding from Cyclic MDS Codes and Expander Graphs
Authors	Netanel Raviv, Itzhak Tamo, Rashish Tandon, Alexandros G. Dimakis
Abstract	Gradient coding is a technique for straggler mitigation in distributed learning. In this paper we design novel gradient codes using tools from classical coding theory, namely, cyclic MDS codes, which compare favorably with existing solutions, both in the applicable range of parameters and in the complexity of the involved algorithms. Second, we introduce an approximate variant of the gradient coding problem, in which we settle for approximate gradient computation instead of the exact one. This approach enables graceful degradation, i.e., the $\ell_2$ error of the approximate gradient is a decreasing function of the number of stragglers. Our main result is that normalized adjacency matrices of expander graphs yield excellent approximate gradient codes, which enable significantly less computation compared to exact gradient coding, and guarantee faster convergence than trivial solutions under standard assumptions. We experimentally test our approach on Amazon EC2, and show that the generalization error of approximate gradient coding is very close to the full gradient while requiring significantly less computation from the workers.
Tasks
Published	2017-07-12
URL	https://arxiv.org/abs/1707.03858v3
PDF	https://arxiv.org/pdf/1707.03858v3.pdf
PWC	https://paperswithcode.com/paper/gradient-coding-from-cyclic-mds-codes-and
Repo
Framework

Neural Multi-Atlas Label Fusion: Application to Cardiac MR Images


Title	Neural Multi-Atlas Label Fusion: Application to Cardiac MR Images
Authors	Heran Yang, Jian Sun, Huibin Li, Lisheng Wang, Zongben Xu
Abstract	Multi-atlas segmentation approach is one of the most widely-used image segmentation techniques in biomedical applications. There are two major challenges in this category of methods, i.e., atlas selection and label fusion. In this paper, we propose a novel multi-atlas segmentation method that formulates multi-atlas segmentation in a deep learning framework for better solving these challenges. The proposed method, dubbed deep fusion net (DFN), is a deep architecture that integrates a feature extraction subnet and a non-local patch-based label fusion (NL-PLF) subnet in a single network. The network parameters are learned by end-to-end training for automatically learning deep features that enable optimal performance in a NL-PLF framework. The learned deep features are further utilized in defining a similarity measure for atlas selection. By evaluating on two public cardiac MR datasets of SATA-13 and LV-09 for left ventricle segmentation, our approach achieved 0.833 in averaged Dice metric (ADM) on SATA-13 dataset and 0.95 in ADM for epicardium segmentation on LV-09 dataset, comparing favorably with the other automatic left ventricle segmentation methods. We also tested our approach on Cardiac Atlas Project (CAP) testing set of MICCAI 2013 SATA Segmentation Challenge, and our method achieved 0.815 in ADM, ranking highest at the time of writing.
Tasks	Semantic Segmentation
Published	2017-09-27
URL	http://arxiv.org/abs/1709.09641v2
PDF	http://arxiv.org/pdf/1709.09641v2.pdf
PWC	https://paperswithcode.com/paper/neural-multi-atlas-label-fusion-application
Repo
Framework