检索结果-内蒙古大学图书馆

Patient risk stratification with time-varying parameters: a multitask learning approach

学校读者我要写书评

暂无评论

The Journal of Machine Learning Research 2016年第1期17卷

作者： Kevin Murphy Bernhard Schölkopf Jenna Wiens John Guttag Eric Horvitz Google MPI for Intelligent Systems Computer Science & Engineering University of Michigan Ann Arbor MI Department of EECS Massachusetts Institute of Technology Cambridge MA Microsoft Research Redmond WA

The proliferation of electronic health records (EHRs) frames opportunities for using machine learning to build models that help healthcare providers improve patient outcomes. However, building useful risk stratification models presents many technical challenges including the large number of factors (both intrinsic and extrinsic) influencing a patient's risk of an adverse outcome and the inherent evolution of that risk over time. We address these challenges in the context of learning a risk stratification model for predicting which patients are at risk of acquiring a Clostridium difficile infection (CDI). We take a novel data-centric approach, leveraging the contents of EHRs from nearly 50,000 hospital admissions. We show how, by adapting techniques from multitask learning, we can learn models for patient risk stratification with unprecedented classification performance. Our model, based on thousands of variables, both time-varying and time-invariant, changes over the course of a patient admission. Applied to a held out set of approximately 25,000 patient admissions, we achieve an area under the receiver operating characteristic curve of 0.81 (95% CI 0.78-0.84). The model has been integrated into the health record system at a large hospital in the US, and can be used to produce daily risk estimates for each inpatient. While more complex than traditional risk stratification methods, the widespread development and use of such data-driven models could ultimately enable cost-effective, targeted prevention strategies that lead to better patient outcomes.

关键词： Clostridium difficile healthcare-associated infections multitask learning risk stratification time-varying coefficients

Low-rank doubly stochastic matrix decomposition for cluster analysis

学校读者我要写书评

暂无评论

The Journal of Machine Learning Research 2016年第1期17卷

作者： Kevin Murphy Bernhard Schölkopf Zhirong Yang Jukka Corander Erkki Oja Google MPI for Intelligent Systems Helsinki Institute of Information Technology University of Helsinki Finland Department of Mathematics and Statistics Helsinki Institute for Information Technology University of Helsinki Finland and Department of Biostatistics University of Oslo Norway Department of Computer Science Aalto University Finland

Cluster analysis by nonnegative low-rank approximations has experienced a remarkable progress in the past decade. However, the majority of such approximation approaches are still restricted to nonnegative matrix factorization (NMF) and suffer from the following two drawbacks: 1) they are unable to produce balanced partitions for large-scale manifold data which are common in real-world clustering tasks; 2) most existing NMF-type clustering methods cannot automatically determine the number of clusters. We propose a new low-rank learning method to address these two problems, which is beyond matrix factorization. Our method approximately decomposes a sparse input similarity in a normalized way and its objective can be used to learn both cluster assignments and the number of clusters. For efficient optimization, we use a relaxed formulation based on Data-Cluster-Data random walk, which is also shown to be equivalent to low-rank factorization of the doubly-stochastically normalized cluster incidence matrix. The probabilistic cluster assignments can thus be learned with a multiplicative majorization-minimization algorithm. Experimental results show that the new method is more accurate both in terms of clustering large-scale manifold data sets and of selecting the number of clusters.

关键词： cluster analysis doubly stochastic matrix manifold multiplicative updates probabilistic relaxation

Hybrid orthogonal projection and estimation (HOPE): a new framework to learn neural networks

学校读者我要写书评

暂无评论

The Journal of Machine Learning Research 2016年第1期17卷

作者： Kevin Murphy Bernhard Schölkopf Shiliang Zhang Hui Jiang Lirong Dai Google MPI for Intelligent Systems National Engineering Laboratory for Speech and Language Information Processing University of Science and Technology of China Hefei Anhui China Department of Electrical Engineering and Computer Science York UniversityToronto OntarioCanada

In this paper, we propose a novel model for high-dimensional data, called the Hybrid Orthogonal Projection and Estimation (HOPE) model, which combines a linear orthogonal projection and a finite mixture model under a unified generative modeling framework. The HOPE model itself can be learned unsupervised from unlabelled data based on the maximum likelihood estimation as well as discriminatively from labelled data. More interestingly, we have shown the proposed HOPE models are closely related to neural networks (NNs) in a sense that each hidden layer can be reformulated as a HOPE model. As a result, the HOPE framework can be used as a novel tool to probe why and how NNs work, more importantly, to learn NNs in either supervised or unsupervised ways. In this work, we have investigated the HOPE framework to learn NNs for several standard tasks, including image recognition on MNIST and speech recognition on TIMIT. Experimental results have shown that the HOPE framework yields significant performance gains over the current state-of-the-art methods in various types of NN learning problems, including unsupervised feature learning, supervised or semi-supervised learning.

关键词： PCA mixture models neural networks orthogonal projection unsupervised learning von Mises-Fisher (vMF) model

Searching for Widespread Events in Large Networked Systems by Cooperative Monitoring

学校读者我要写书评

暂无评论

Searching for Widespread Events in Large Networked Systems b...

International Conference on Network Protocols

作者： Zhiping Cai Min Chen Shigang Chen Yan Qiao College of Computer National University of Defense Technology Changsha Hunan Department of Computer & Information Science & Engineering University of Florida Gainesville FL USA Google Inc. CA USA

Searching for widespread events in large networks is a fundamental function that underlies many important applications of distributed anomaly detection, traffic measurement, online data mining, etc. This function can be performed by a cooperative monitoring system consisting of a central coordinator and a number of monitors that are deployed at a set of vantage points. We formulate a network primitive function, called multi-monitor joint detection, which is to find the common events observed by all or a given subset of monitors during each measurement period. It is a challenging problem because large-scale cooperative monitoring can generate tremendous communication overhead. Therefore, it is critical to design a solution for multi-monitor joint detection which controls communication overhead to a low level. We thoroughly examine existing techniques that may be applied, and identify their performance limitations. We then propose two new techniques, called combinable filters and progressive filtering, which address the performance limitations from different angles. We formally prove the correctness of our new solutions based on a probabilistic joint detection model. Numerical evaluation shows that our best solution achieves an overhead reduction in the range of 63% to 91% over the Bloom filter solution under various simulation settings when the number of monitors is 10 or more.

关键词： Monitoring Probabilistic logic Numerical models Distributed databases Protocols google Ports (computers)

Estimating diffusion networks: recovery conditions, sample complexity & soft-thresholding algorithm

学校读者我要写书评

暂无评论

The Journal of Machine Learning Research 2016年第1期17卷

作者： Kevin Murphy Bernhard Schölkopf Manuel Gomez-Rodriguez Le Song Hadi Daneshmand Google MPI for Intelligent Systems Tübingen Germany College of Computing MPI for Software Systems Kaiserslautern Germany College of Computing Georgia Institute of Technology Atlanta GA Computer Science Department Zürich Switzerland

Information spreads across social and technological networks, but often the network structures are hidden from us and we only observe the traces left by the diffusion processes, called cascades. Can we recover the hidden network structures from these observed cascades? What kind of cascades and how many cascades do we need? Are there some network structures which are more difficult than others to recover? Can we design efficient inference algorithms with provable guarantees?Despite the increasing availability of cascade data and methods for inferring networks from these data, a thorough theoretical understanding of the above questions remains largely unexplored in the literature. In this paper, we investigate the network structure inference problem for a general family of continuous-time diffusion models using an l1- regularized likelihood maximization framework. We show that, as long as the cascade sampling process satisfies a natural incoherence condition, our framework can recover the correct network structure with high probability if we observe O(d3logN) cascades, where d is the maximum number of parents of a node and N is the total number of nodes. Moreover, we develop a simple and efficient soft-thresholding network inference algorithm which demonstrate the match between our theoretical prediction and empirical results. In practice, this new algorithm also outperforms other alternatives in terms of the accuracy of recovering hidden diffusion networks.

关键词：

Measuring dependence powerfully and equitably

学校读者我要写书评

暂无评论

The Journal of Machine Learning Research 2016年第1期17卷

作者： Kevin Murphy Bernhard Schölkopf Yakir A. Reshef David N. Reshef Hilary K. Finucane Pardis C. Sabeti Michael Mitzenmacher Google MPI for Intelligent Systems School of Engineering and Applied Sciences Harvard University Cambridge MA Department of Electrical Engineering and Computer Science Massachusetts Institute of Technology Cambridge MA Department of Mathematics Massachusetts Institute of Technology Cambridge MA Department of Organismic and Evolutionary Biology Harvard University Cambridge MA

Given a high-dimensional data set, we often wish to find the strongest relationships within it. A common strategy is to evaluate a measure of dependence on every variable pair and retain the highest-scoring pairs for follow-up. This strategy works well if the statistic used (a) has good power to detect non-trivial relationships, and (b) is equitable, meaning that for some measure of noise it assigns similar scores to equally noisy relationships regardless of relationship type (e.g., linear, exponential, periodic). In this paper, we define and theoretically characterize two new statistics that together yield an efficient approach for obtaining both power and equitability. To do this, we first introduce a new population measure of dependence and show three equivalent ways that it can be viewed, including as a canonical "smoothing" of mutual information. We then introduce an efficiently computable consistent estimator of our population measure of dependence, and we empirically establish its equitability on a large class of noisy functional relationships. This new statistic has better bias/variance properties and better runtime complexity than a previous heuristic approach. Next, we derive a second, related statistic whose computation is a trivial side-product of our algorithm and whose goal is powerful independence testing rather than equitability. We prove that this statistic yields a consistent independence test and show in simulations that the test has good power against independence. Taken together, our results suggest that these two statistics are a valuable pair of tools for exploratory data analysis.

关键词： equitability maximal information coefficient mutual information statistical power total information coefficient

Approximate shorteSt descending paths

学校读者我要写书评

暂无评论

Approximate shorteSt descending paths

作者： Cheng, Siu-Wing Jin, Jiongxin Department of Computer Science and Engineering Hong Kong University of Science and Technology Hong Kong Google Inc Seattle WA 98103 United States

We present an approximation algorithm for the shorteSt descending path problem. Given a source s and a deStination t on a terrain, a shorteSt descending path from s to t is a path of minimum Euclidean length on the terrain subject to the conStraint that the height decreases monotonically as we traverse that path from s to t. Given any Ε ∈ (0, 1), our algorithm returns in O(n4 log(n/Ε)) time a descending path of length at moSt 1 + Ε times the optimum. This is the firSt algorithm whose running time is polynomial in n and log(1/Ε) and independent of the terrain geometry. © by SIAM. Unauthorized reproduction of this CPticle is prohibited.

关键词： Approximation algorithms

Multiple output regression with latent noise

学校读者我要写书评

暂无评论

The Journal of Machine Learning Research 2016年第1期17卷

作者： Kevin Murphy Bernhard Schölkopf Jussi Gillberg Pekka Marttinen Matti Pirinen Antti J. Kangas Pasi Soininen Mehreen Ali Aki S. Havulinna Marjo-Riitta Järvelin Mika Ala-Korpela Samuel Kaski Google MPI for Intelligent Systems Helsinki Institute for Information Technology Department of Computer Science Aalto University Aalto Finland Institute for Molecular Medicine Finland University of Helsinki Finland Computational Medicine Faculty of Medicine University of Oulu & Biocenter Oulu Oulu Finland Department of Health National Institute for Health and Welfare Helsinki Finland Department of Epidemiology and Biostatistics MRC-PHE Centre for Environment & Health School of Public Health Imperial College London UK

In high-dimensional data, structured noise caused by observed and unobserved factors affecting multiple target variables simultaneously, imposes a serious challenge for modeling, by masking the often weak signal. Therefore, (1) explaining away the structured noise in multiple-output regression is of paramount importance. Additionally, (2) assumptions about the correlation structure of the regression weights are needed. We note that both can be formulated in a natural way in a latent variable model, in which both the interesting signal and the noise are mediated through the same latent factors. Under this assumption, the signal model then borrows strength from the noise model by encouraging similar effects on correlated targets. We introduce a hyperparameter for the latent signal-to-noise ratio which turns out to be important for modelling weak signals, and an ordered infinite-dimensional shrinkage prior that resolves the rotational unidentifiability in reduced-rank regression models. Simulations and prediction experiments with metabolite, gene expression, FMRI measurement, and macroeconomic time series data show that our model equals or exceeds the state-of-the-art performance and, in particular, outperforms the standard approach of assuming independent noise and signal models.

关键词： Bayesian reduced-rank regression latent signal-to-noise ratio latent variable models multiple-output regression nonparametric Bayes shrinkage priors structured noise weak effects

Exploiting group recommendation functions for flexible preferences

学校读者我要写书评

暂无评论

Exploiting group recommendation functions for flexible prefe...

30th IEEE International Conference on Data Engineering, ICDE 2014

作者： Basu Roy, Senjuti Thirumuruganathan, Saravanan Amer-Yahia, Sihem Das, Gautam Yu, Cong Institute of Technology University of Washington Tacoma United States Computer Science Department University of Texas at Arlington United States CNRS-LIG United States Google Research United States

ISBN: (纸本)9781479925544

We examine the problem of enabling the flexibility of updating one's preferences in group recommendation. In our setting, any group member can provide a vector of preferences that, in addition to past preferences and other group members' preferences, will be accounted for in computing group recommendation. This functionality is essential in many group recommendation applications, such as travel planning, online games, book clubs, or strategic voting, as it has been previously shown that user preferences may vary depending on mood, context, and company (i.e., other people in the group). Preferences are enforced in an feedback box that replaces preferences provided by the users by a potentially different feedback vector that is better suited for maximizing the individual satisfaction when computing the group recommendation. The feedback box interacts with a traditional recommendation box that implements a group consensus semantics in the form of Aggregated Voting or Least Misery, two popular aggregation functions for group recommendation. We develop efficient algorithms to compute robust group recommendations that are appropriate in situations where users have changing preferences. Our extensive empirical study on real world data-sets validates our findings. © 2014 IEEE.

关键词： Semantics