Showing papers by "Stanford University published in 2011"

Distributed Optimization and Statistical Learning Via the Alternating Direction Method of Multipliers

Abstract: Many problems of recent interest in statistics and machine learning can be posed in the framework of convex optimization. Due to the explosion in size and complexity of modern datasets, it is increasingly important to be able to solve problems with a very large number of features or training examples. As a result, both the decentralized collection or storage of these datasets as well as accompanying distributed solution methods are either necessary or at least highly desirable. In this review, we argue that the alternating direction method of multipliers is well suited to distributed convex optimization, and in particular to large-scale problems arising in statistics, machine learning, and related areas. The method was developed in the 1970s, with roots in the 1950s, and is equivalent or closely related to many other algorithms, such as dual decomposition, the method of multipliers, Douglas–Rachford splitting, Spingarn's method of partial inverses, Dykstra's alternating projections, Bregman iterative algorithms for l1 problems, proximal methods, and others. After briefly surveying the theory and history of the algorithm, we discuss applications to a wide variety of statistical and machine learning problems of recent interest, including the lasso, sparse logistic regression, basis pursuit, covariance selection, support vector machines, and many others. We also discuss general distributed optimization, extensions to the nonconvex setting, and efficient implementation, including some details on distributed MPI and Hadoop MapReduce implementations.

Topological insulators and superconductors

Abstract: Topological insulators are new states of quantum matter which cannot be adiabatically connected to conventional insulators and semiconductors. They are characterized by a full insulating gap in the bulk and gapless edge or surface states which are protected by time-reversal symmetry. These topological materials have been theoretically predicted and experimentally observed in a variety of systems, including HgTe quantum wells, BiSb alloys, and Bi2Te3 and Bi2Se3 crystals. Theoretical models, materials properties, and experimental results on two-dimensional and three-dimensional topological insulators are reviewed, and both the topological band theory and the topological field theory are discussed. Topological superconductors have a full pairing gap in the bulk and gapless surface states consisting of Majorana fermions. The theory of topological superconductors is reviewed, in close analogy to the theory of topological insulators.

Robust principal component analysis

Abstract: This article is about a curious phenomenon. Suppose we have a data matrix, which is the superposition of a low-rank component and a sparse component. Can we recover each component individuallyq We prove that under some suitable assumptions, it is possible to recover both the low-rank and the sparse components exactly by solving a very convenient convex program called Principal Component Pursuit; among all feasible decompositions, simply minimize a weighted combination of the nuclear norm and of the e1 norm. This suggests the possibility of a principled approach to robust principal component analysis since our methodology and results assert that one can recover the principal components of a data matrix even though a positive fraction of its entries are arbitrarily corrupted. This extends to the situation where a fraction of the entries are missing as well. We discuss an algorithm for solving this optimization problem, and present applications in the area of video surveillance, where our methodology allows for the detection of objects in a cluttered background, and in the area of face recognition, where it offers a principled way of removing shadows and specularities in images of faces.

Integrated genomic analyses of ovarian carcinoma

Abstract: A catalogue of molecular aberrations that cause ovarian cancer is critical for developing and deploying therapies that will improve patients' lives. The Cancer Genome Atlas project has analysed messenger RNA expression, microRNA expression, promoter methylation and DNA copy number in 489 high-grade serous ovarian adenocarcinomas and the DNA sequences of exons from coding genes in 316 of these tumours. Here we report that high-grade serous ovarian cancer is characterized by TP53 mutations in almost all tumours (96%); low prevalence but statistically recurrent somatic mutations in nine further genes including NF1, BRCA1, BRCA2, RB1 and CDK12; 113 significant focal DNA copy number aberrations; and promoter methylation events involving 168 genes. Analyses delineated four ovarian cancer transcriptional subtypes, three microRNA subtypes, four promoter methylation subtypes and a transcriptional signature associated with survival duration, and shed new light on the impact that tumours with BRCA1/2 (BRCA1 or BRCA2) and CCNE1 aberrations have on survival. Pathway analyses suggested that homologous recombination is defective in about half of the tumours analysed, and that NOTCH and FOXM1 signalling are involved in serous ovarian cancer pathophysiology.

Reading Digits in Natural Images with Unsupervised Feature Learning

Abstract: Detecting and reading text from natural images is a hard computer vision task that is central to a variety of emerging applications. Related problems like document character recognition have been widely studied by computer vision and machine learning researchers and are virtually solved for practical applications like reading handwritten digits. Reliably recognizing characters in more complex scenes like photographs, however, is far more difficult: the best existing methods lag well behind human performance on the same tasks. In this paper we attack the problem of recognizing digits in a real application using unsupervised feature learning methods: reading house numbers from street level photos. To this end, we introduce a new benchmark dataset for research use containing over 600,000 labeled digits cropped from Street View images. We then demonstrate the difficulty of recognizing these digits when the problem is approached with hand-designed features. Finally, we employ variants of two recently proposed unsupervised feature learning methods and find that they are convincingly superior on our benchmarks.

Co3O4 nanocrystals on graphene as a synergistic catalyst for oxygen reduction reaction

Abstract: Catalysts for oxygen reduction and evolution reactions are at the heart of key renewable-energy technologies including fuel cells and water splitting. Despite tremendous efforts, developing oxygen electrode catalysts with high activity at low cost remains a great challenge. Here, we report a hybrid material consisting of Co₃O₄ nanocrystals grown on reduced graphene oxide as a high-performance bi-functional catalyst for the oxygen reduction reaction (ORR) and oxygen evolution reaction (OER). Although Co₃O₄ or graphene oxide alone has little catalytic activity, their hybrid exhibits an unexpected, surprisingly high ORR activity that is further enhanced by nitrogen doping of graphene. The Co₃O₄/N-doped graphene hybrid exhibits similar catalytic activity but superior stability to Pt in alkaline solutions. The same hybrid is also highly active for OER, making it a high-performance non-precious metal-based bi-catalyst for both ORR and OER. The unusual catalytic activity arises from synergetic chemical coupling effects between Co₃O₄ and graphene.

2011 Compendium of Physical Activities: a second update of codes and MET values.

Abstract: Purpose:The Compendium of Physical Activities was developed to enhance the comparability of results across studies using self-report physical activity (PA) and is used to quantify the energy cost of a wide variety of PA. We provide the second update of the Compendium, called the 2011 Compend

A statistical explanation of MaxEnt for ecologists

Abstract: MaxEnt is a program for modelling species distributions from presence-only species records. This paper is written for ecologists and describes the MaxEnt model from a statistical perspective, making explicit links between the structure of the model, decisions required in producing a modelled distribution, and knowledge about the species and the data that might affect those decisions. To begin we discuss the characteristics of presence-only data, highlighting implications for modelling distributions. We particularly focus on the problems of sample bias and lack of information on species prevalence. The keystone of the paper is a new statistical explanation of MaxEnt which shows that the model minimizes the relative entropy between two probability densities (one estimated from the presence data and one, from the landscape) defined in covariate space. For many users, this viewpoint is likely to be a more accessible way to understand the model than previous ones that rely on machine learning concepts. We then step through a detailed explanation of MaxEnt describing key components (e.g. covariates and features, and definition of the landscape extent), the mechanics of model fitting (e.g. feature selection, constraints and regularization) and outputs. Using case studies for a Banksia species native to south-west Australia and a riverine fish, we fit models and interpret them, exploring why certain choices affect the result and what this means. The fish example illustrates use of the model with vector data for linear river segments rather than raster (gridded) data. Appropriate treatments for survey bias, unprojected data, locally restricted species, and predicting to environments outside the range of the training data are demonstrated, and new capabilities discussed. Online appendices include additional details of the model and the mathematical links between previous explanations and this one, example code and data, and further information on the case studies.

Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials

Abstract: Funnel plots, and tests for funnel plot asymmetry, have been widely used to examine bias in the results of meta-analyses. Funnel plot asymmetry should not be equated with publication bias, because it has a number of other possible causes. This article describes how to interpret funnel plot asymmetry, recommends appropriate tests, and explains the implications for choice of meta-analysis model

MoS2 Nanoparticles Grown on Graphene: An Advanced Catalyst for the Hydrogen Evolution Reaction

Abstract: Advanced materials for electrocatalytic and photoelectrochemical water splitting are central to the area of renewable energy. In this work, we developed a selective solvothermal synthesis of MoS2 nanoparticles on reduced graphene oxide (RGO) sheets suspended in solution. The resulting MoS2/RGO hybrid material possessed nanoscopic few-layer MoS2 structures with an abundance of exposed edges stacked onto graphene, in strong contrast to large aggregated MoS2 particles grown freely in solution without GO. The MoS2/RGO hybrid exhibited superior electrocatalytic activity in the hydrogen evolution reaction (HER) relative to other MoS2 catalysts. A Tafel slope of ∼41 mV/decade was measured for MoS2 catalysts in the HER for the first time; this exceeds by far the activity of previous MoS2 catalysts and results from the abundance of catalytic edge sites on the MoS2 nanoparticles and the excellent electrical coupling to the underlying graphene network. The ∼41 mV/decade Tafel slope suggested the Volmer–Heyrovsky mec...

Learning Word Vectors for Sentiment Analysis

Abstract: Unsupervised vector-based approaches to semantics can model rich lexical meanings, but they largely fail to capture sentiment information that is central to many word meanings and important for a wide range of NLP tasks. We present a model that uses a mix of unsupervised and supervised techniques to learn word vectors capturing semantic term--document information as well as rich sentiment content. The proposed model can leverage both continuous and multi-dimensional sentiment information as well as non-sentiment annotations. We instantiate the model to utilize the document-level sentiment polarity annotations present in many online documents (e.g. star ratings). We evaluate the model using small, widely used sentiment and subjectivity corpora and find it out-performs several previously introduced methods for sentiment classification. We also introduce a large dataset of movie reviews to serve as a more robust benchmark for work in this area.

Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials

Abstract: Most state-of-the-art techniques for multi-class image segmentation and labeling use conditional random fields defined over pixels or image regions. While region-level models often feature dense pairwise connectivity, pixel-level models are considerably larger and have only permitted sparse graph structures. In this paper, we consider fully connected CRF models defined on the complete set of pixels in an image. The resulting graphs have billions of edges, making traditional inference algorithms impractical. Our main contribution is a highly efficient approximate inference algorithm for fully connected CRF models in which the pairwise edge potentials are defined by a linear combination of Gaussian kernels. Our experiments demonstrate that dense connectivity at the pixel level substantially improves segmentation and labeling accuracy.

Climate Trends and Global Crop Production Since 1980

Abstract: Efforts to anticipate how climate change will affect future food availability can benefit from understanding the impacts of changes to date. We found that in the cropping regions and growing seasons of most countries, with the important exception of the United States, temperature trends from 1980 to 2008 exceeded one standard deviation of historic year-to-year variability. Models that link yields of the four largest commodity crops to weather indicate that global maize and wheat production declined by 3.8 and 5.5%, respectively, relative to a counterfactual without climate trends. For soybeans and rice, winners and losers largely balanced out. Climate trends were large enough in some countries to offset a significant portion of the increases in average yields that arose from technology, carbon dioxide fertilization, and other factors.

Regression shrinkage and selection via the lasso: a retrospective

Abstract: Summary. In the paper I give a brief review of the basic idea and some history and then discuss some developments since the original paper on regression shrinkage and selection via the lasso.

MatchIt: Nonparametric Preprocessing for Parametric Causal Inference

Abstract: MatchIt implements the suggestions of Ho, Imai, King, and Stuart (2007) for improving parametric statistical models by preprocessing data with nonparametric matching methods. MatchIt implements a wide range of sophisticated matching methods, making it possible to greatly reduce the dependence of causal inferences on hard-to-justify, but commonly made, statistical modeling assumptions. The software also easily fits into existing research practices since, after preprocessing data with MatchIt , researchers can use whatever parametric model they would have used without MatchIt , but produce inferences with substantially more robustness and less sensitivity to modeling assumptions. MatchIt is an R program, and also works seamlessly with Zelig .

Universality in Oxygen Evolution Electrocatalysis on Oxide Surfaces

Abstract: Trends in electrocatalytic activity of the oxygen evolution reaction (OER) are investigated on the basis of a large database of HO* and HOO* adsorption energies on oxide surfaces. The theoretical overpotential was calculated by applying standard density functional theory in combination with the computational standard hydrogen electrode (SHE) model. We showed that by the discovery of a universal scaling relation between the adsorption energies of HOO* vs HO*, it is possible to analyze the reaction free energy diagrams of all the oxides in a general way. This gave rise to an activity volcano that was the same for a wide variety of oxide catalyst materials and a universal descriptor for the oxygen evolution activity, which suggests a fundamental limitation on the maximum oxygen evolution activity of planar oxide catalysts.

Friendship and mobility: user movement in location-based social networks

Abstract: Even though human movement and mobility patterns have a high degree of freedom and variation, they also exhibit structural patterns due to geographic and social constraints. Using cell phone location data, as well as data from two online location-based social networks, we aim to understand what basic laws govern human motion and dynamics. We find that humans experience a combination of periodic movement that is geographically limited and seemingly random jumps correlated with their social networks. Short-ranged travel is periodic both spatially and temporally and not effected by the social network structure, while long-distance travel is more influenced by social network ties. We show that social relationships can explain about 10% to 30% of all human movement, while periodic behavior explains 50% to 70%. Based on our findings, we develop a model of human mobility that combines periodic short range movements with travel due to the social network structure. We show that our model reliably predicts the locations and dynamics of future human movement and gives an order of magnitude better performance than present models of human mobility.

Skin-like pressure and strain sensors based on transparent elastic films of carbon nanotubes

Abstract: Transparent films of carbon nanotubes can accommodate strains of up to 150% and demonstrate conductivities as high as 2,200 S cm−1 in the stretched state.

Multimodal Deep Learning

Abstract: Deep networks have been successfully applied to unsupervised feature learning for single modalities (e.g., text, images or audio). In this work, we propose a novel application of deep networks to learn features over multiple modalities. We present a series of tasks for multimodal learning and show how to train deep networks that learn features to address these tasks. In particular, we demonstrate cross modality feature learning, where better features for one modality (e.g., video) can be learned if multiple modalities (e.g., audio and video) are present at feature learning time. Furthermore, we show how to learn a shared representation between modalities and evaluate it on a unique task, where the classifier is trained with audio-only data but tested with video-only data and vice-versa. Our models are validated on the CUAVE and AVLetters datasets on audio-visual speech classification, demonstrating best published visual speech classification on AVLetters and effective shared representation learning.

Crystal structure of the β2 adrenergic receptor-Gs protein complex.

Abstract: G protein-coupled receptors (GPCRs) are responsible for the majority of cellular responses to hormones and neurotransmitters as well as the senses of sight, olfaction and taste. The paradigm of GPCR signalling is the activation of a heterotrimeric GTP binding protein (G protein) by an agonist-occupied receptor. The b2 adrenergic receptor (b2AR) activation of Gs, the stimulatory G protein for adenylyl cyclase, has long been a model system for GPCR signalling. Here we present the crystal structure of the active state ternary complex composed of agonist-occupied monomericb2AR and nucleotide-free Gs heterotrimer. The principal interactions between the b2AR and Gs involve the amino- and carboxy-terminal a-helices of Gs, with conformational changes propagating to the nucleotide-binding pocket. The

D³ Data-Driven Documents

Abstract: Data-Driven Documents (D3) is a novel representation-transparent approach to visualization for the web Rather than hide the underlying scenegraph within a toolkit-specific abstraction, D3 enables direct inspection and manipulation of a native representation: the standard document object model (DOM) With D3, designers selectively bind input data to arbitrary document elements, applying dynamic transforms to both generate and modify content We show how representational transparency improves expressiveness and better integrates with developer tools than prior approaches, while offering comparable notational efficiency and retaining powerful declarative components Immediate evaluation of operators further simplifies debugging and allows iterative development Additionally, we demonstrate how D3 transforms naturally enable animation and interaction with dramatic performance improvements over intermediate representations

2017 Guidelines of the American Thyroid Association for the Diagnosis and Management of Thyroid Disease During Pregnancy and the Postpartum

Abstract: Background: Thyroid disease in pregnancy is a common clinical problem. Since the guidelines for the management of these disorders by the American Thyroid Association (ATA) were first published in 2...

Autophagy proteins regulate innate immune responses by inhibiting the release of mitochondrial DNA mediated by the NALP3 inflammasome

Abstract: Autophagy, a cellular process for organelle and protein turnover, regulates innate immune responses. Here we demonstrate that depletion of the autophagic proteins LC3B and beclin 1 enhanced the activation of caspase-1 and secretion of interleukin 1β (IL-1β) and IL-18. Depletion of autophagic proteins promoted the accumulation of dysfunctional mitochondria and cytosolic translocation of mitochondrial DNA (mtDNA) in response to lipopolysaccharide (LPS) and ATP in macrophages. Release of mtDNA into the cytosol depended on the NALP3 inflammasome and mitochondrial reactive oxygen species (ROS). Cytosolic mtDNA contributed to the secretion of IL-1β and IL-18 in response to LPS and ATP. LC3B-deficient mice produced more caspase-1-dependent cytokines in two sepsis models and were susceptible to LPS-induced mortality. Our study suggests that autophagic proteins regulate NALP3-dependent inflammation by preserving mitochondrial integrity.

Single-Cell Mass Cytometry of Differential Immune and Drug Responses Across a Human Hematopoietic Continuum

Abstract: Flow cytometry is an essential tool for dissecting the functional complexity of hematopoiesis. We used single-cell "mass cytometry" to examine healthy human bone marrow, measuring 34 parameters simultaneously in single cells (binding of 31 antibodies, viability, DNA content, and relative cell size). The signaling behavior of cell subsets spanning a defined hematopoietic hierarchy was monitored with 18 simultaneous markers of functional signaling states perturbed by a set of ex vivo stimuli and inhibitors. The data set allowed for an algorithmically driven assembly of related cell types defined by surface antigen expression, providing a superimposable map of cell signaling responses in combination with drug inhibition. Visualized in this manner, the analysis revealed previously unappreciated instances of both precise signaling responses that were bounded within conventionally defined cell subsets and more continuous phosphorylation responses that crossed cell population boundaries in unexpected manners yet tracked closely with cellular phenotype. Collectively, such single-cell analyses provide system-wide views of immune signaling in healthy human hematopoiesis, against which drug action and disease can be compared for mechanistic studies and pharmacologic intervention.

An analysis of single-layer networks in unsupervised feature learning

Abstract: A great deal of research has focused on algorithms for learning features from unlabeled data. Indeed, much progress has been made on benchmark datasets like NORB and CIFAR by employing increasingly complex unsupervised learning algorithms and deep models. In this paper, however, we show that several simple factors, such as the number of hidden nodes in the model, may be more important to achieving high performance than the learning algorithm or the depth of the model. Specifically, we will apply several othe-shelf feature learning algorithms (sparse auto-encoders, sparse RBMs, K-means clustering, and Gaussian mixtures) to CIFAR, NORB, and STL datasets using only singlelayer networks. We then present a detailed analysis of the eect of changes in the model setup: the receptive field size, number of hidden nodes (features), the step-size (“stride”) between extracted features, and the eect of whitening. Our results show that large numbers of hidden nodes and dense feature extraction are critical to achieving high performance—so critical, in fact, that when these parameters are pushed to their limits, we achieve state-of-the-art performance on both CIFAR-10 and NORB using only a single layer of features. More surprisingly, our best performance is based on K-means clustering, which is extremely fast, has no hyperparameters to tune beyond the model structure itself, and is very easy to implement. Despite the simplicity of our system, we achieve accuracy beyond all previously published results on the CIFAR-10 and NORB datasets (79.6% and 97.2% respectively).

Global land use change, economic globalization, and the looming land scarcity

Abstract: A central challenge for sustainability is how to preserve forest ecosystems and the services that they provide us while enhancing food production. This challenge for developing countries confronts the force of economic globalization, which seeks cropland that is shrinking in availability and triggers deforestation. Four mechanisms—the displacement, rebound, cascade, and remittance effects—that are amplified by economic globalization accelerate land conversion. A few developing countries have managed a land use transition over the recent decades that simultaneously increased their forest cover and agricultural production. These countries have relied on various mixes of agricultural intensification, land use zoning, forest protection, increased reliance on imported food and wood products, the creation of off-farm jobs, foreign capital investments, and remittances. Sound policies and innovations can therefore reconcile forest preservation with food production. Globalization can be harnessed to increase land use efficiency rather than leading to uncontrolled land use expansion. To do so, land systems should be understood and modeled as open systems with large flows of goods, people, and capital that connect local land use with global-scale factors.

CANDELS: The Cosmic Assembly Near-infrared Deep Extragalactic Legacy Survey

Abstract: The Cosmic Assembly Near-IR Deep Extragalactic Legacy Survey (CANDELS) is designed to document the first third of galactic evolution, from z approx. 8 - 1.5. It will image > 250,000 distant galaxies using three separate cameras on the Hubble Space Tele8cope, from the mid-UV to near-IR, and will find and measure Type Ia supernovae beyond z > 1.5 to test their accuracy as standard candles for cosmology. Five premier multi-wavelength sky regions are selected, each with extensive ancillary data. The use of five widely separated fields mitigates cosmic variance and yields statistically robust and complete samples of galaxies down to a stellar mass of 10(exp 9) solar mass to z approx. 2, reaching the knee of the UV luminosity function of galaxies to z approx. 8. The survey covers approximately 800 square arc minutes and is divided into two parts. The CANDELS/Deep survey (5(sigma) point-source limit H =27.7mag) covers approx. 125 square arcminutes within GOODS-N and GOODS-S. The CANDELS/Wide survey includes GOODS and three additional fields (EGS, COSMOS, and UDS) and covers the full area to a 50(sigma) point-source limit of H ? or approx. = 27.0 mag. Together with the Hubble Ultradeep Fields, the strategy creates a three-tiered "wedding cake" approach that has proven efficient for extragalactic surveys. Data from the survey are non-proprietary and are useful for a wide variety of science investigations. In this paper, we describe the basic motivations for the survey, the CANDELS team science goals and the resulting observational requirements, the field selection and geometry, and the observing design.

A unique chromatin signature uncovers early developmental enhancers in humans

Abstract: Cell-fate transitions involve the integration of genomic information encoded by regulatory elements, such as enhancers, with the cellular environment. However, identification of genomic sequences that control human embryonic development represents a formidable challenge. Here we show that in human embryonic stem cells (hESCs), unique chromatin signatures identify two distinct classes of genomic elements, both of which are marked by the presence of chromatin regulators p300 and BRG1, monomethylation of histone H3 at lysine 4 (H3K4me1), and low nucleosomal density. In addition, elements of the first class are distinguished by the acetylation of histone H3 at lysine 27 (H3K27ac), overlap with previously characterized hESC enhancers, and are located proximally to genes expressed in hESCs and the epiblast. In contrast, elements of the second class, which we term 'poised enhancers', are distinguished by the absence of H3K27ac, enrichment of histone H3 lysine 27 trimethylation (H3K27me3), and are linked to genes inactive in hESCs and instead are involved in orchestrating early steps in embryogenesis, such as gastrulation, mesoderm formation and neurulation. Consistent with the poised identity, during differentiation of hESCs to neuroepithelium, a neuroectoderm-specific subset of poised enhancers acquires a chromatin signature associated with active enhancers. When assayed in zebrafish embryos, poised enhancers are able to direct cell-type and stage-specific expression characteristic of their proximal developmental gene, even in the absence of sequence conservation in the fish genome. Our data demonstrate that early developmental enhancers are epigenetically pre-marked in hESCs and indicate an unappreciated role of H3K27me3 at distal regulatory elements. Moreover, the wealth of new regulatory sequences identified here provides an invaluable resource for studies and isolation of transient, rare cell populations representing early stages of human embryogenesis.