October 17, 2019

2991 words 15 mins read

Paper Group ANR 869

Machine Learning for Exam Triage. Regional Multi-scale Approach for Visually Pleasing Explanations of Deep Neural Networks. Knockoffs for the mass: new feature importance statistics with false discovery guarantees. Assessment of iris recognition reliability for eyes affected by ocular pathologies. Analytic heuristics for a fast DSC-MRI. Non-local O …

Machine Learning for Exam Triage


Title	Machine Learning for Exam Triage
Authors	Xinyu Guan, Jessica Lee, Peter Wu, Yue Wu
Abstract	In this project, we extend the state-of-the-art CheXNet (Rajpurkar et al. [2017]) by making use of the additional non-image features in the dataset. Our model produced better AUROC scores than the original CheXNet.
Tasks
Published	2018-04-30
URL	http://arxiv.org/abs/1805.00503v1
PDF	http://arxiv.org/pdf/1805.00503v1.pdf
PWC	https://paperswithcode.com/paper/machine-learning-for-exam-triage
Repo
Framework

Regional Multi-scale Approach for Visually Pleasing Explanations of Deep Neural Networks


Title	Regional Multi-scale Approach for Visually Pleasing Explanations of Deep Neural Networks
Authors	Dasom Seo, Kanghan Oh, Il-Seok Oh
Abstract	Recently, many methods to interpret and visualize deep neural network predictions have been proposed and significant progress has been made. However, a more class-discriminative and visually pleasing explanation is required. Thus, this paper proposes a region-based approach that estimates feature importance in terms of appropriately segmented regions. By fusing the saliency maps generated from multi-scale segmentations, a more class-discriminative and visually pleasing map is obtained. We incorporate this regional multi-scale concept into a prediction difference method that is model-agnostic. An input image is segmented in several scales using the super-pixel method, and exclusion of a region is simulated by sampling a normal distribution constructed using the boundary prior. The experimental results demonstrate that the regional multi-scale method produces much more class-discriminative and visually pleasing saliency maps.
Tasks	Feature Importance
Published	2018-07-31
URL	http://arxiv.org/abs/1807.11720v2
PDF	http://arxiv.org/pdf/1807.11720v2.pdf
PWC	https://paperswithcode.com/paper/regional-multi-scale-approach-for-visually
Repo
Framework

Knockoffs for the mass: new feature importance statistics with false discovery guarantees


Title	Knockoffs for the mass: new feature importance statistics with false discovery guarantees
Authors	Jaime Roquero Gimenez, Amirata Ghorbani, James Zou
Abstract	An important problem in machine learning and statistics is to identify features that causally affect the outcome. This is often impossible to do from purely observational data, and a natural relaxation is to identify features that are correlated with the outcome even conditioned on all other observed features. For example, we want to identify that smoking really is correlated with cancer conditioned on demographics. The knockoff procedure is a recent breakthrough in statistics that, in theory, can identify truly correlated features while guaranteeing that the false discovery is limited. The idea is to create synthetic data – knockoffs – that captures correlations amongst the features. However there are substantial computational and practical challenges to generating and using knockoffs. This paper makes several key advances that enable knockoff application to be more efficient and powerful. We develop an efficient algorithm to generate valid knockoffs from Bayesian Networks. Then we systematically evaluate knockoff test statistics and develop new statistics with improved power. The paper combines new mathematical guarantees with systematic experiments on real and synthetic data.
Tasks	Feature Importance
Published	2018-07-17
URL	https://arxiv.org/abs/1807.06214v2
PDF	https://arxiv.org/pdf/1807.06214v2.pdf
PWC	https://paperswithcode.com/paper/knockoffs-for-the-mass-new-feature-importance
Repo
Framework

Assessment of iris recognition reliability for eyes affected by ocular pathologies


Title	Assessment of iris recognition reliability for eyes affected by ocular pathologies
Authors	Mateusz Trokielewicz, Adam Czajka, Piotr Maciejewicz
Abstract	This paper presents an analysis of how the iris recognition is impacted by eye diseases and an appropriate dataset comprising 2996 iris images of 230 distinct eyes (including 184 illness-affected eyes representing more than 20 different eye conditions). The images were collected in near infrared and visible light during a routine ophthalmological practice. The experimental study shows four valuable results. First, the enrollment process is highly sensitive to those eye conditions that make the iris obstructed or introduce geometrical distortions. Second, even those conditions that do not produce visible changes to the iris structure may increase the dissimilarity among samples of the same eyes. Third, eye conditions affecting iris geometry, its tissue structure or producing obstructions significantly decrease the iris recognition reliability. Fourth, for eyes afflicted by a disease, the most prominent effect of the disease on iris recognition is to cause segmentation errors. To our knowledge this is the first database of iris images for disease-affected eyes made publicly available to researchers, and the most comprehensive study of what we can expect when the iris recognition is deployed for non-healthy eyes.
Tasks	Iris Recognition
Published	2018-09-01
URL	http://arxiv.org/abs/1809.00206v1
PDF	http://arxiv.org/pdf/1809.00206v1.pdf
PWC	https://paperswithcode.com/paper/assessment-of-iris-recognition-reliability
Repo
Framework

Analytic heuristics for a fast DSC-MRI


Title	Analytic heuristics for a fast DSC-MRI
Authors	Marco Virgulin, Marco Castellaro, Enrico Grisan, Fabio Marcuzzi
Abstract	In this paper we propose a deterministic approach for the reconstruction of Dynamic Susceptibility Contrast magnetic resonance imaging data and compare it with the compressed sensing solution existing in the literature for the same problem. Our study is based on the mathematical analysis of the problem, which is computationally intractable because of its non polynomial complexity, but suggests simple heuristics that perform quite well. We give results on real images and on artificial phantoms with added noise.
Tasks
Published	2018-12-11
URL	http://arxiv.org/abs/1812.04303v1
PDF	http://arxiv.org/pdf/1812.04303v1.pdf
PWC	https://paperswithcode.com/paper/analytic-heuristics-for-a-fast-dsc-mri
Repo
Framework

Non-local Operational Anisotropic Diffusion Filter


Title	Non-local Operational Anisotropic Diffusion Filter
Authors	Fábio A. M. Cappabianco, Petrus P. C. E. da Silva
Abstract	High-frequency noise is present in several modalities of medical images. It originates from the acquisition process and may be related to the scanner configurations, the scanned body, or to other external factors. This way, prospective filters are an important tool to improve the image quality. In this paper, we propose a non-local weighted operational anisotropic diffusion filter and evaluate its effect on magnetic resonance images and on kV/CBCT radiotherapy images. We also provide a detailed analysis of non-local parameter settings. Results show that the new filter enhances previous local implementations and has potential application in radiotherapy treatments.
Tasks
Published	2018-12-11
URL	http://arxiv.org/abs/1812.04708v1
PDF	http://arxiv.org/pdf/1812.04708v1.pdf
PWC	https://paperswithcode.com/paper/non-local-operational-anisotropic-diffusion
Repo
Framework

Towards the Development of a Rule-based Drought Early Warning Expert Systems using Indigenous Knowledge


Title	Towards the Development of a Rule-based Drought Early Warning Expert Systems using Indigenous Knowledge
Authors	A. K. Akanbi, M. Masinde
Abstract	Drought forecasting and prediction is a complicated process due to the complexity and scalability of the environmental parameters involved. Hence, it required a high level of expertise to predict. In this paper, we describe the research and development of a rule-based drought early warning expert systems (RB-DEWES) for forecasting drought using local indigenous knowledge obtained from domain experts. The system generates inference by using rule set and provides drought advisory information with attributed certainty factor (CF) based on the user’s input. The system is believed to be the first expert system for drought forecasting to use local indigenous knowledge on drought. The architecture and components such as knowledge base, JESS inference engine and model base of the system and their functions are presented.
Tasks
Published	2018-09-19
URL	http://arxiv.org/abs/1809.08101v1
PDF	http://arxiv.org/pdf/1809.08101v1.pdf
PWC	https://paperswithcode.com/paper/towards-the-development-of-a-rule-based
Repo
Framework

Stochastic Approximation for Risk-aware Markov Decision Processes


Title	Stochastic Approximation for Risk-aware Markov Decision Processes
Authors	Wenjie Huang, William B. Haskell
Abstract	We develop a stochastic approximation-type algorithm to solve finite state/action, infinite-horizon, risk-aware Markov decision processes. Our algorithm has two loops. The inner loop computes the risk by solving a stochastic saddle-point problem. The outer loop performs $Q$-learning to compute an optimal risk-aware policy. Several widely investigated risk measures (e.g. conditional value-at-risk, optimized certainty equivalent, and absolute semi-deviation) are covered by our algorithm. Almost sure convergence and the convergence rate of the algorithm are established. For an error tolerance $\epsilon>0$ for the optimal $Q$-value estimation gap and learning rate $k\in(1/2,,1]$, the overall convergence rate of our algorithm is $\Omega((\ln(1/\delta\epsilon)/\epsilon^{2})^{1/k}+(\ln(1/\epsilon))^{1/(1-k)})$ with probability at least $1-\delta$.
Tasks	Q-Learning
Published	2018-05-11
URL	https://arxiv.org/abs/1805.04238v4
PDF	https://arxiv.org/pdf/1805.04238v4.pdf
PWC	https://paperswithcode.com/paper/stochastic-approximation-for-risk-aware
Repo
Framework

The Convergence of Stochastic Gradient Descent in Asynchronous Shared Memory


Title	The Convergence of Stochastic Gradient Descent in Asynchronous Shared Memory
Authors	Dan Alistarh, Christopher De Sa, Nikola Konstantinov
Abstract	Stochastic Gradient Descent (SGD) is a fundamental algorithm in machine learning, representing the optimization backbone for training several classic models, from regression to neural networks. Given the recent practical focus on distributed machine learning, significant work has been dedicated to the convergence properties of this algorithm under the inconsistent and noisy updates arising from execution in a distributed environment. However, surprisingly, the convergence properties of this classic algorithm in the standard shared-memory model are still not well-understood. In this work, we address this gap, and provide new convergence bounds for lock-free concurrent stochastic gradient descent, executing in the classic asynchronous shared memory model, against a strong adaptive adversary. Our results give improved upper and lower bounds on the “price of asynchrony” when executing the fundamental SGD algorithm in a concurrent setting. They show that this classic optimization tool can converge faster and with a wider range of parameters than previously known under asynchronous iterations. At the same time, we exhibit a fundamental trade-off between the maximum delay in the system and the rate at which SGD can converge, which governs the set of parameters under which this algorithm can still work efficiently.
Tasks
Published	2018-03-23
URL	http://arxiv.org/abs/1803.08841v2
PDF	http://arxiv.org/pdf/1803.08841v2.pdf
PWC	https://paperswithcode.com/paper/the-convergence-of-stochastic-gradient
Repo
Framework

Geometric Understanding of Deep Learning


Title	Geometric Understanding of Deep Learning
Authors	Na Lei, Zhongxuan Luo, Shing-Tung Yau, David Xianfeng Gu
Abstract	Deep learning is the mainstream technique for many machine learning tasks, including image recognition, machine translation, speech recognition, and so on. It has outperformed conventional methods in various fields and achieved great successes. Unfortunately, the understanding on how it works remains unclear. It has the central importance to lay down the theoretic foundation for deep learning. In this work, we give a geometric view to understand deep learning: we show that the fundamental principle attributing to the success is the manifold structure in data, namely natural high dimensional data concentrates close to a low-dimensional manifold, deep learning learns the manifold and the probability distribution on it. We further introduce the concepts of rectified linear complexity for deep neural network measuring its learning capability, rectified linear complexity of an embedding manifold describing the difficulty to be learned. Then we show for any deep neural network with fixed architecture, there exists a manifold that cannot be learned by the network. Finally, we propose to apply optimal mass transportation theory to control the probability distribution in the latent space.
Tasks	Machine Translation, Speech Recognition
Published	2018-05-26
URL	http://arxiv.org/abs/1805.10451v2
PDF	http://arxiv.org/pdf/1805.10451v2.pdf
PWC	https://paperswithcode.com/paper/geometric-understanding-of-deep-learning
Repo
Framework

Enriched Interpretation


Title	Enriched Interpretation
Authors	Robert E. Kent
Abstract	The theory introduced, presented and developed in this paper, is concerned with an enriched extension of the theory of Rough Sets pioneered by Zdzislaw Pawlak. The enrichment discussed here is in the sense of valuated categories as developed by F.W. Lawvere. This paper relates Rough Sets to an abstraction of the theory of Fuzzy Sets pioneered by Lotfi Zadeh, and provides a natural foundation for “soft computation”. To paraphrase Lotfi Zadeh, the impetus for the transition from a hard theory to a soft theory derives from the fact that both the generality of a theory and its applicability to real-world problems are substantially enhanced by replacing various hard concepts with their soft counterparts. Here we discuss the corresponding enriched notions for indiscernibility, subsets, upper/lower approximations, and rough sets. Throughout, we indicate linkages with the theory of Formal Concept Analysis pioneered by Rudolf Wille. We pay particular attention to the all-important notion of a “linguistic variable” - developing its enriched extension, comparing it with the notion of conceptual scale from Formal Concept Analysis, and discussing the pragmatic issues of its creation and use in the interpretation of data. These pragmatic issues are exemplified by the discovery, conceptual analysis, interpretation, and categorization of networked information resources in WAVE, the Web Analysis and Visualization Environment currently being developed for the management and interpretation of the universe of resource information distributed over the World-Wide Web.
Tasks
Published	2018-10-20
URL	http://arxiv.org/abs/1810.08831v1
PDF	http://arxiv.org/pdf/1810.08831v1.pdf
PWC	https://paperswithcode.com/paper/enriched-interpretation
Repo
Framework

Learning Particle Dynamics for Manipulating Rigid Bodies, Deformable Objects, and Fluids


Title	Learning Particle Dynamics for Manipulating Rigid Bodies, Deformable Objects, and Fluids
Authors	Yunzhu Li, Jiajun Wu, Russ Tedrake, Joshua B. Tenenbaum, Antonio Torralba
Abstract	Real-life control tasks involve matters of various substances—rigid or soft bodies, liquid, gas—each with distinct physical behaviors. This poses challenges to traditional rigid-body physics engines. Particle-based simulators have been developed to model the dynamics of these complex scenes; however, relying on approximation techniques, their simulation often deviates from real-world physics, especially in the long term. In this paper, we propose to learn a particle-based simulator for complex control tasks. Combining learning with particle-based systems brings in two major benefits: first, the learned simulator, just like other particle-based systems, acts widely on objects of different materials; second, the particle-based representation poses strong inductive bias for learning: particles of the same type have the same dynamics within. This enables the model to quickly adapt to new environments of unknown dynamics within a few observations. We demonstrate robots achieving complex manipulation tasks using the learned simulator, such as manipulating fluids and deformable foam, with experiments both in simulation and in the real world. Our study helps lay the foundation for robot learning of dynamic scenes with particle-based representations.
Tasks
Published	2018-10-03
URL	http://arxiv.org/abs/1810.01566v2
PDF	http://arxiv.org/pdf/1810.01566v2.pdf
PWC	https://paperswithcode.com/paper/learning-particle-dynamics-for-manipulating
Repo
Framework

Understanding Self-Paced Learning under Concave Conjugacy Theory


Title	Understanding Self-Paced Learning under Concave Conjugacy Theory
Authors	Shiqi Liu, Zilu Ma, Deyu Meng
Abstract	By simulating the easy-to-hard learning manners of humans/animals, the learning regimes called curriculum learning~(CL) and self-paced learning~(SPL) have been recently investigated and invoked broad interests. However, the intrinsic mechanism for analyzing why such learning regimes can work has not been comprehensively investigated. To this issue, this paper proposes a concave conjugacy theory for looking into the insight of CL/SPL. Specifically, by using this theory, we prove the equivalence of the SPL regime and a latent concave objective, which is closely related to the known non-convex regularized penalty widely used in statistics and machine learning. Beyond the previous theory for explaining CL/SPL insights, this new theoretical framework on one hand facilitates two direct approaches for designing new SPL models for certain tasks, and on the other hand can help conduct the latent objective of self-paced curriculum learning, which is the advanced version of both CL/SPL and possess advantages of both learning regimes to a certain extent. This further facilitates a theoretical understanding for SPCL, instead of only CL/SPL as conventional. Under this theory, we attempt to attain intrinsic latent objectives of two curriculum forms, the partial order and group curriculums, which easily follow the theoretical understanding of the corresponding SPCL regimes.
Tasks
Published	2018-05-21
URL	http://arxiv.org/abs/1805.08096v1
PDF	http://arxiv.org/pdf/1805.08096v1.pdf
PWC	https://paperswithcode.com/paper/understanding-self-paced-learning-under
Repo
Framework

SUSTain: Scalable Unsupervised Scoring for Tensors and its Application to Phenotyping


Title	SUSTain: Scalable Unsupervised Scoring for Tensors and its Application to Phenotyping
Authors	Ioakeim Perros, Evangelos E. Papalexakis, Haesun Park, Richard Vuduc, Xiaowei Yan, Christopher Defilippi, Walter F. Stewart, Jimeng Sun
Abstract	This paper presents a new method, which we call SUSTain, that extends real-valued matrix and tensor factorizations to data where values are integers. Such data are common when the values correspond to event counts or ordinal measures. The conventional approach is to treat integer data as real, and then apply real-valued factorizations. However, doing so fails to preserve important characteristics of the original data, thereby making it hard to interpret the results. Instead, our approach extracts factor values from integer datasets as scores that are constrained to take values from a small integer set. These scores are easy to interpret: a score of zero indicates no feature contribution and higher scores indicate distinct levels of feature importance. At its core, SUSTain relies on: a) a problem partitioning into integer-constrained subproblems, so that they can be optimally solved in an efficient manner; and b) organizing the order of the subproblems’ solution, to promote reuse of shared intermediate results. We propose two variants, SUSTain_M and SUSTain_T, to handle both matrix and tensor inputs, respectively. We evaluate SUSTain against several state-of-the-art baselines on both synthetic and real Electronic Health Record (EHR) datasets. Comparing to those baselines, SUSTain shows either significantly better fit or orders of magnitude speedups that achieve a comparable fit (up to 425X faster). We apply SUSTain to EHR datasets to extract patient phenotypes (i.e., clinically meaningful patient clusters). Furthermore, 87% of them were validated as clinically meaningful phenotypes related to heart failure by a cardiologist.
Tasks	Feature Importance
Published	2018-03-14
URL	http://arxiv.org/abs/1803.05473v1
PDF	http://arxiv.org/pdf/1803.05473v1.pdf
PWC	https://paperswithcode.com/paper/sustain-scalable-unsupervised-scoring-for
Repo
Framework

Beyond Inferring Class Representatives: User-Level Privacy Leakage From Federated Learning


Title	Beyond Inferring Class Representatives: User-Level Privacy Leakage From Federated Learning
Authors	Zhibo Wang, Mengkai Song, Zhifei Zhang, Yang Song, Qian Wang, Hairong Qi
Abstract	Federated learning, i.e., a mobile edge computing framework for deep learning, is a recent advance in privacy-preserving machine learning, where the model is trained in a decentralized manner by the clients, i.e., data curators, preventing the server from directly accessing those private data from the clients. This learning mechanism significantly challenges the attack from the server side. Although the state-of-the-art attacking techniques that incorporated the advance of Generative adversarial networks (GANs) could construct class representatives of the global data distribution among all clients, it is still challenging to distinguishably attack a specific client (i.e., user-level privacy leakage), which is a stronger privacy threat to precisely recover the private data from a specific client. This paper gives the first attempt to explore user-level privacy leakage against the federated learning by the attack from a malicious server. We propose a framework incorporating GAN with a multi-task discriminator, which simultaneously discriminates category, reality, and client identity of input samples. The novel discrimination on client identity enables the generator to recover user specified private data. Unlike existing works that tend to interfere the training process of the federated learning, the proposed method works “invisibly” on the server side. The experimental results demonstrate the effectiveness of the proposed attacking approach and the superior to the state-of-the-art.
Tasks
Published	2018-12-03
URL	http://arxiv.org/abs/1812.00535v3
PDF	http://arxiv.org/pdf/1812.00535v3.pdf
PWC	https://paperswithcode.com/paper/beyond-inferring-class-representatives-user
Repo
Framework