arXiv subjects

Shu Yang

Publications and source records attributed to Shu Yang.

At least 253 records · Page 14Linked to original sources

arXiv2019

Semiparametric estimation of structural failure time model in continuous-time processes

Structural failure time models are causal models for estimating the effect of time-varying treatments on a survival outcome. G-estimation and artificial censoring have been proposed to estimate the model parameters in the presence of time-dependent confounding and administrative censoring. However, most of existing methods require manually preprocessing data into regularly spaced data, which may invalidate the subsequent causal analysis. Moreover, the computation and inference are challenging due to the non-smoothness of artificial censoring. We propose a class of continuous-time structural failure time models, which respects the continuous time nature of the underlying data processes. Under a martingale condition of no unmeasured confounding, we show that the model parameters are identifiable from potentially infinite estimating equations. Using the semiparametric efficiency theory, we derive the first semiparametric doubly robust estimators, in the sense that the estimators are consistent if either the treatment process model or the failure time model is correctly specified, but not necessarily both. Moreover, we propose using inverse probability of censoring weighting to deal with dependent censoring. In contrast to artificial censoring, our weighting strategy does not introduce non-smoothness in estimation and ensures that the resampling methods can be used to make inference.

stat.ME↗

arXiv2019

Causal inference with confounders missing not at random

It is important to draw causal inference from observational studies, which, however, becomes challenging if the confounders have missing values. Generally, causal effects are not identifiable if the confounders are missing not at random. We propose a novel framework to nonparametrically identify causal effects with confounders subject to an outcome-independent missingness, that is, the missing data mechanism is independent of the outcome, given the treatment and possibly missing confounders. We then propose a nonparametric two-stage least squares estimator and a parametric estimator for causal effects.

stat.ME↗

arXiv2018

Penetrating the Fog: the Path to Efficient CNN Models

With the increasing demand to deploy convolutional neural networks (CNNs) on mobile platforms, the sparse kernel approach was proposed, which could save more parameters than the standard convolution while maintaining accuracy. However, despite the great potential, no prior research has pointed out how to craft an sparse kernel design with such potential (i.e., effective design), and all prior works just adopt simple combinations of existing sparse kernels such as group convolution. Meanwhile due to the large design space it is also impossible to try all combinations of existing sparse kernels. In this paper, we are the first in the field to consider how to craft an effective sparse kernel design by eliminating the large design space. Specifically, we present a sparse kernel scheme to illustrate how to reduce the space from three aspects. First, in terms of composition we remove designs composed of repeated layers. Second, to remove designs with large accuracy degradation, we find an unified property named information field behind various sparse kernel designs, which could directly indicate the final accuracy. Last, we remove designs in two cases where a better parameter efficiency could be achieved. Additionally, we provide detailed efficiency analysis on the final four designs in our scheme. Experimental results validate the idea of our scheme by showing that our scheme is able to find designs which are more efficient in using parameters and computation with similar or higher accuracy.

cs.CV↗

arXiv2018

Domain-Adversarial Multi-Task Framework for Novel Therapeutic Property Prediction of Compounds

With the rapid development of high-throughput technologies, parallel acquisition of large-scale drug-informatics data provides huge opportunities to improve pharmaceutical research and development. One significant application is the purpose prediction of small molecule compounds, aiming to specify therapeutic properties of extensive purpose-unknown compounds and to repurpose novel therapeutic properties of FDA-approved drugs. Such problem is very challenging since compound attributes contain heterogeneous data with various feature patterns such as drug fingerprint, drug physicochemical property, drug perturbation gene expression. Moreover, there is complex nonlinear dependency among heterogeneous data. In this paper, we propose a novel domain-adversarial multi-task framework for integrating shared knowledge from multiple domains. The framework utilizes the adversarial strategy to effectively learn target representations and models their nonlinear dependency. Experiments on two real-world datasets illustrate that the performance of our approach obtains an obvious improvement over competitive baselines. The novel therapeutic properties of purpose-unknown compounds we predicted are mostly reported or brought to the clinics. Furthermore, our framework can integrate various attributes beyond the three domains examined here and can be applied in the industry for screening the purpose of huge amounts of as yet unidentified compounds. Source codes of this paper are available on Github.

cs.LG↗

arXiv2018

SECS: Efficient Deep Stream Processing via Class Skew Dichotomy

Despite that accelerating convolutional neural network (CNN) receives an increasing research focus, the save on resource consumption always comes with a decrease in accuracy. To both increase accuracy and decrease resource consumption, we explore an environment information, called class skew, which is easily available and exists widely in daily life. Since the class skew may switch as time goes, we bring up probability layer to utilize class skew without any overhead during the runtime. Further, we observe class skew dichotomy that some class skew may appear frequently in the future, called hot class skew, and others will never appear again or appear seldom, called cold class skew. Inspired by techniques from source code optimization, two modes, i.e., interpretation and compilation, are proposed. The interpretation mode pursues efficient adaption during runtime for cold class skew and the compilation mode aggressively optimize on hot ones for more efficient deployment in the future. Aggressive optimization is processed by class-specific pruning and provides extra benefit. Finally, we design a systematic framework, SECS, to dynamically detect class skew, processing interpretation and compilation, as well as select the most accurate architectures under the runtime resource budget. Extensive evaluations show that SECS can realize end-to-end classification speedups by a factor of 3x to 11x relative to state-of-the-art convolutional neural networks, at a higher accuracy.

cs.CV↗

arXiv2018

Integration of survey data and big observational data for finite population inference using mass imputation

Multiple data sources are becoming increasingly available for statistical analyses in the era of big data. As an important example in finite-population inference, we consider an imputation approach to combining a probability sample with big observational data. Unlike the usual imputation for missing data analysis, we create imputed values for the whole elements in the probability sample. Such mass imputation is attractive in the context of survey data integration (Kim and Rao, 2012). We extend mass imputation as a tool for data integration of survey data and big non-survey data. The mass imputation methods and their statistical properties are presented. The matching estimator of Rivers (2007) is also covered as a special case. Variance estimation with mass-imputed data is discussed. The simulation results demonstrate the proposed estimators outperform existing competitors in terms of robustness and efficiency.

stat.ME↗

arXiv2018

Shaping nanoparticle fingerprints at the interface of cholesteric droplets

The ordering of nanoparticles into predetermined configurations is of importance to the design of advanced technologies. In this work, we moderate the surface anchoring against the bulk elasticity of liquid crystals to dynamically shape nanoparticle assemblies at a fluid interface. By tuning the degree of nanoparticle hydrophobicity with surfactants that alter the molecular anchoring of liquid crystals, we pattern nanoparticles at the interface of cholesteric liquid crystal emulsions. Adjusting the particle hydrophobicity more finely further modifies the rigidity of assemblies. We establish that patterns are tunable by varying both surfactant and chiral dopant concentrations. Since particle assembly occurs at the interface with the desired structures exposed to the surrounding phase, we demonstrate that particles can be readily crosslinked and manipulated, forming structures that retain their shape under external perturbations. This study establishes the templating of nanomaterials into reconfigurable arrangements. Interfacial assembly is tempered by elastic patterns that arise from the geometric frustration of confined cholesterics. This work serves as a basis for creating materials with chemical heterogeneity and with linear, periodic structures, essential for optical and energy applications.

cond-mat.soft↗

arXiv2018

Elastocapillary driven assembly of particles at free-standing smectic-A films

Colloidal particles at complex fluid interfaces and within films assemble to form ordered structures with high degrees of symmetry via interactions that include capillarity, elasticity, and other fields like electrostatic charge. Here we study microparticle interactions within free-standing smectic-A films, in which the elasticity arising from the director field distortion and capillary interactions arising from interface deformation compete to direct the assembly of motile particles. New colloidal assemblies and patterns, ranging from 1D chains to 2D aggregates, sensitive to the initial wetting conditions of particles at the smectic film, are reported. This work paves the way to exploiting LC interfaces as a means to direct spontaneously formed, reconfigurable, and optically active materials.

cond-mat.soft↗

arXiv2018

Predictive mean matching imputation in survey sampling

Predictive mean matching imputation is popular for handling item nonresponse in survey sampling. In this article, we study the asymptotic properties of the predictive mean matching estimator of the population mean. For variance estimation, the conventional bootstrap inference for matching estimators with fixed matches has been shown to be invalid due to the nonsmoothness nature of the matching estimator. We propose asymptotically valid replication variance estimation. The key strategy is to construct replicates of the estimator directly based on linear terms, instead of individual records of variables. Extension to nearest neighbor imputation is also discussed. A simulation study confirms that the new procedure provides valid variance estimation.

stat.ME↗

arXiv2017

Making Faces: Universal Inverse Design of Surfaces with Thin Nematic Elastomer Sheets

Programmable shape-shifting materials can take different physical forms to achieve multifunctionality in a dynamic and controllable manner. Although morphing a shape from 2D to 3D via programmed inhomogeneous local deformations has been demonstrated in various ways, the inverse problem -- programming a sheet to take an arbitrary desired 3D shape -- is much harder yet critical to realize specific functions. Here, we address this inverse problem in thin liquid crystal elastomer (LCE) sheets, where the shape is preprogrammed by precise and local control of the molecular orientation of the liquid crystal monomers. We show how blueprints for arbitrary surface geometries as well as local extrinsic curvatures can be generated using approximate numerical methods. Backed by faithfully alignable and rapidly lockable LCE chemistry, we precisely embed our designs in LCE sheets using advanced top-down microfabrication techniques. We thus successfully produce flat sheets that, upon thermal activation, take an arbitrary desired shape, such as a face. The general design principles presented here for creating an arbitrary 3D shape will allow for exploration of unmet needs in flexible electronics, metamaterials, aerospace and medical devices, and more.

cond-mat.soft↗

arXiv2017

Nearest neighbor imputation for general parameter estimation in survey sampling

Nearest neighbor imputation is popular for handling item nonresponse in survey sampling. In this article, we study the asymptotic properties of the nearest neighbor imputation estimator for general population parameters, including population means, proportions and quantiles. For variance estimation, the conventional bootstrap inference for matching estimators with fixed number of matches has been shown to be invalid due to the nonsmoothness nature of the matching estimator. We propose asymptotically valid replication variance estimation. The key strategy is to construct replicates of the estimator directly based on linear terms, instead of individual records of variables. A simulation study confirms that the new procedure provides valid variance estimation.

stat.ME↗

arXiv2017

Propensity score weighting for causal inference with clustered data

Propensity score weighting is a tool for causal inference to adjust for measured confounders in observational studies. In practice, data often present complex structures, such as clustering, which make propensity score modeling and estimation challenging. In addition, for clustered data, there may be unmeasured cluster-specific variables that are related to both the treatment assignment and the outcome. When such unmeasured cluster-specific confounders exist and are omitted in the propensity score model, the subsequent propensity score adjustment may be biased. In this article, we propose a calibration technique for propensity score estimation under the latent ignorable treatment assignment mechanism, i.e., the treatment-outcome relationship is unconfounded given the observed covariates and the latent cluster effects. We then provide a consistent propensity score weighting estimator of the average treatment effect when the propensity score and outcome follow generalized linear mixed effects models. The proposed propensity score weighting estimator is attractive, because it does not require specification of functional forms of the propensity score and outcome models, and therefore is robust to model misspecification. The proposed weighting method can be combined with sampling weights for an integrated solution to handle confounding and sampling designs for causal inference with clustered survey data. In simulation studies, we show that the proposed estimator is superior to other competitors. We estimate the effect of School Body Mass Index Screening on prevalence of overweight and obesity for elementary schools in Pennsylvania.

stat.ME↗

arXiv2016

Propensity score weighting for causal inference with multi-stage clustered data

Propensity score weighting is a tool for causal inference to adjust for measured confounders. Survey data are often collected under complex sampling designs such as multistage cluster sampling, which presents challenges for propensity score modeling and estimation. In addition, for clustered data, there may also be unobserved cluster effects related to both the treatment and the outcome. When such unmeasured confounders exist and are omitted in the propensity score model, the subsequent propensity score adjustment will be biased. We propose a calibrated propensity score weighting adjustment for multi-stage clustered data in the presence of unmeasured cluster-level confounders. The propensity score is calibrated to balance design-weighted covariate distributions and cluster effects between treatment groups. In particular, we consider a growing number of calibration constraints increasing with the number of clusters, which is necessary for removing asymptotic bias that is associated with the unobserved cluster-level confounders. We show that our estimator is robust in the sense that the estimator is consistent without correct specification of the propensity score model. We extend the results to the multiple treatments case. In simulation studies we show that the proposed estimator is superior to other competitors. We estimate the effect of School Body Mass Index Screening on prevalence of overweight and obesity for elementary schools in Pennsylvania.

stat.ME↗

arXiv2015

Propensity Score Matching and Subclassification in Observational Studies with Multi-level Treatments

In this paper, we develop new methods for estimating average treatment effects in observational studies, focusing on settings with more than two treatment levels under unconfoundedness given pre-treatment variables. We emphasize subclassification and matching methods which have been found to be effective in the binary treatment literature and which are among the most popular methods in that setting. Whereas the literature has suggested that these particular propensity-based methods do not naturally extend to the multi-level treatment case, we show, using the concept of weak unconfoundedness, that adjusting for or matching on a scalar function of the pre-treatment variables removes all biases associated with observed pre-treatment variables. We apply the proposed methods to an analysis of the effect of treatments for fibromyalgia. We also carry out a simulation study to assess the finite sample performance of the methods relative to previously proposed methods.

stat.ME↗

arXiv2015

Doubly Robust Goodness-of-Fit Test of Coarse Structural Nested Mean Models with Application to Initiating combination antiretroviral treatment in HIV-Positive Patients

Coarse Structural Nested Mean Models (SNMMs) provide useful tools to estimate treatment effects from longitudinal observational data with time-dependent confounders. Coarse SNMMs lead to a large class of estimators,within which an optimal estimator can be derived under the conditions of well-specified models for the treatment effect, for treatment initiation, and for nuisance regression outcomes (Lok & Griner, 2015). The key assumption lies in a well-specified model for the treatment effect; however, there is no existing guidance to specify the treatment effect model, and model misspecification leads to biased estimators, preventing valid inference. To test whether the treatment effect model matches the data well, we derive a goodness-of-fit (GOF) test procedure based on overidentification restrictions tests (Sargan, 1958; Hansen, 1982). We show that our GOF statistic is doubly-robust in the sense that with a correct treatment effect model, if either the treatment initiation model or the nuisance regression outcome model is correctly specified, the GOF statistic has a Chi-squared limiting distribution with degrees of freedom equal to the number of overidentification restrictions. We demonstrate the empirical relevance of our methods using simulation designs based on an actual dataset. In addition, we apply the GOF test procedure to study how the initiation time of highly active antiretroviral treatment (HAART) after infection predicts the one-year treatment effect in HIV-positive patients with acute and early infection.

stat.ME↗

arXiv2015

Semiparametric estimation of spectral density function for irregular spatial data

Estimation of the covariance structure of spatial processes is of fundamental importance in spatial statistics. In the literature, several non-parametric and semi-parametric methods have been developed to estimate the covariance structure based on the spectral representation of covariance functions. However,they either ignore the high frequency properties of the spectral density, which are essential to determine the performance of interpolation procedures such as Kriging, or lack of theoretical justification. We propose a new semi-parametric method to estimate spectral densities of isotropic spatial processes with irregular observations. The spectral density function at low frequencies is estimated using smoothing spline, while a parametric model is used for the spectral density at high frequencies, and the parameters are estimated by a method-of-moment approach based on empirical variograms at small lags. We derive the asymptotic bounds for bias and variance of the proposed estimator. The simulation study shows that our method outperforms the existing non-parametric estimator by several performance criteria.

stat.ME↗

arXiv2015

A note on multiple imputation for method of moments estimation

Multiple imputation is a popular imputation method for general purpose estimation. Rubin(1987) provided an easily applicable formula for the variance estimation of multiple imputation. However, the validity of the multiple imputation inference requires the congeniality condition of Meng(1994), which is not necessarily satisfied for method of moments estimation. This paper presents the asymptotic bias of Rubin's variance estimator when the method of moments estimator is used as a complete-sample estimator in the multiple imputation procedure. A new variance estimator based on over-imputation is proposed to provide asymptotically valid inference for method of moments estimation.

stat.ME↗

arXiv2015

Fractional Imputation in Survey Sampling: A Comparative Review

Fractional imputation (FI) is a relatively new method of imputation for handling item nonresponse in survey sampling. In FI, several imputed values with their fractional weights are created for each missing item. Each fractional weight represents the conditional probability of the imputed value given the observed data, and the parameters in the conditional probabilities are often computed by an iterative method such as EM algorithm. The underlying model for FI can be fully parametric, semiparametric, or nonparametric, depending on plausibility of assumptions and the data structure. In this paper, we give an overview of FI, introduce key ideas and methods to readers who are new to the FI literature, and highlight some new development. We also provide guidance on practical implementation of FI and valid inferential tools after imputation. We demonstrate the empirical performance of FI with respect to multiple imputation using a pseudo finite population generated from a sample in Monthly Retail Trade Survey in US Census Bureau.

stat.ME↗