arXiv subjects

Jing Lei

Publications and source records attributed to Jing Lei.

71 records · Page 4Linked to original sources

arXiv2016

Learning with Differential Privacy: Stability, Learnability and the Sufficiency and Necessity of ERM Principle

While machine learning has proven to be a powerful data-driven solution to many real-life problems, its use in sensitive domains has been limited due to privacy concerns. A popular approach known as **differential privacy** offers provable privacy guarantees, but it is often observed in practice that it could substantially hamper learning accuracy. In this paper we study the learnability (whether a problem can be learned by any algorithm) under Vapnik's general learning setting with differential privacy constraint, and reveal some intricate relationships between privacy, stability and learnability. In particular, we show that a problem is privately learnable **if an only if** there is a private algorithm that asymptotically minimizes the empirical risk (AERM). In contrast, for non-private learning AERM alone is not sufficient for learnability. This result suggests that when searching for private learning algorithms, we can restrict the search to algorithms that are AERM. In light of this, we propose a conceptual procedure that always finds a universally consistent algorithm whenever the problem is learnable under privacy constraint. We also propose a generic and practical algorithm and show that under very general conditions it privately learns a wide class of learning problems. Lastly, we extend some of the results to the more practical $(ε,δ)$-differential privacy and establish the existence of a phase-transition on the class of problems that are approximately privately learnable with respect to how small $δ$ needs to be.

stat.ML↗

arXiv2016

A Minimax Theory for Adaptive Data Analysis

In adaptive data analysis, the user makes a sequence of queries on the data, where at each step the choice of query may depend on the results in previous steps. The releases are often randomized in order to reduce overfitting for such adaptively chosen queries. In this paper, we propose a minimax framework for adaptive data analysis. Assuming Gaussianity of queries, we establish the first sharp minimax lower bound on the squared error in the order of $O(\frac{\sqrt{k}σ^2}{n})$, where $k$ is the number of queries asked, and $σ^2/n$ is the ordinary signal-to-noise ratio for a single query. Our lower bound is based on the construction of an approximately least favorable adversary who picks a sequence of queries that are most likely to be affected by overfitting. This approximately least favorable adversary uses only one level of adaptivity, suggesting that the minimax risk for 1-step adaptivity with k-1 initial releases and that for $k$-step adaptivity are on the same order. The key technical component of the lower bound proof is a reduction to finding the convoluting distribution that optimally obfuscates the sign of a Gaussian signal. Our lower bound construction also reveals a transparent and elementary proof of the matching upper bound as an alternative approach to Russo and Zou (2015), who used information-theoretic tools to provide the same upper bound. We believe that the proposed framework opens up opportunities to obtain theoretical insights for many other settings of adaptive data analysis, which would extend the idea to more practical realms.

stat.ML↗

arXiv2016

A goodness-of-fit test for stochastic block models

The stochastic block model is a popular tool for studying community structures in network data. We develop a goodness-of-fit test for the stochastic block model. The test statistic is based on the largest singular value of a residual matrix obtained by subtracting the estimated block mean effect from the adjacency matrix. Asymptotic null distribution is obtained using recent advances in random matrix theory. The test is proved to have full power against alternative models with finer structures. These results naturally lead to a consistent sequential testing estimate of the number of communities.

math.ST↗

arXiv2015

Network assisted analysis to reveal the genetic basis of autism

While studies show that autism is highly heritable, the nature of the genetic basis of this disorder remains illusive. Based on the idea that highly correlated genes are functionally interrelated and more likely to affect risk, we develop a novel statistical tool to find more potentially autism risk genes by combining the genetic association scores with gene co-expression in specific brain regions and periods of development. The gene dependence network is estimated using a novel partial neighborhood selection (PNS) algorithm, where node specific properties are incorporated into network estimation for improved statistical and computational efficiency. Then we adopt a hidden Markov random field (HMRF) model to combine the estimated network and the genetic association scores in a systematic manner. The proposed modeling framework can be naturally extended to incorporate additional structural information concerning the dependence between genes. Using currently available genetic association data from whole exome sequencing studies and brain gene expression levels, the proposed algorithm successfully identified 333 genes that plausibly affect autism risk.

stat.ME↗

arXiv2015

Network Cross-Validation for Determining the Number of Communities in Network Data

The stochastic block model and its variants have been a popular tool in analyzing large network data with community structures. In this paper we develop an efficient network cross-validation (NCV) approach to determine the number of communities, as well as to choose between the regular stochastic block model and the degree corrected block model. The proposed NCV method is based on a block-wise node-pair splitting technique, combined with an integrated step of community recovery using sub-blocks of the adjacency matrix. We prove that the probability of under selection vanishes as the number of node increases, under mild conditions satisfied by a wide range of popular community recovery algorithms. The solid performance of our method is also demonstrated in extensive simulations and a data example.

stat.ME↗

arXiv2015

Sparsistency and agnostic inference in sparse PCA

The presence of a sparse "truth" has been a constant assumption in the theoretical analysis of sparse PCA and is often implicit in its methodological development. This naturally raises questions about the properties of sparse PCA methods and how they depend on the assumption of sparsity. Under what conditions can the relevant variables be selected consistently if the truth is assumed to be sparse? What can be said about the results of sparse PCA without assuming a sparse and unique truth? We answer these questions by investigating the properties of the recently proposed Fantope projection and selection (FPS) method in the high-dimensional setting. Our results provide general sufficient conditions for sparsistency of the FPS estimator. These conditions are weak and can hold in situations where other estimators are known to fail. On the other hand, without assuming sparsity or identifiability, we show that FPS provides a sparse, linear dimension-reducing transformation that is close to the best possible in terms of maximizing the predictive covariance.

math.ST↗

arXiv2015

Localized Functional Principal Component Analysis

We propose localized functional principal component analysis (LFPCA), looking for orthogonal basis functions with localized support regions that explain most of the variability of a random process. The LFPCA is formulated as a convex optimization problem through a novel Deflated Fantope Localization method and is implemented through an efficient algorithm to obtain the global optimum. We prove that the proposed LFPCA converges to the original FPCA when the tuning parameters are chosen appropriately. Simulation shows that the proposed LFPCA with tuning parameters chosen by cross validation can almost perfectly recover the true eigenfunctions and significantly improve the estimation accuracy when the eigenfunctions are truly supported on some subdomains. In the scenario that the original eigenfunctions are not localized, the proposed LFPCA also serves as a nice tool in finding orthogonal basis functions that balance between interpretability and the capability of explaining variability of the data. The analyses of a country mortality data and a growth curve data reveal interesting features that cannot be found by standard FPCA methods.

stat.ME↗

arXiv2014

Consistency of spectral clustering in stochastic block models

We analyze the performance of spectral clustering for community extraction in stochastic block models. We show that, under mild conditions, spectral clustering applied to the adjacency matrix of the network can consistently recover hidden communities even when the order of the maximum expected degree is as small as $\log n$, with $n$ the number of nodes. This result applies to some popular polynomial time spectral clustering algorithms and is further extended to degree corrected stochastic block models using a spherical $k$-median spectral clustering method. A key component of our analysis is a combinatorial bound on the spectrum of binary random matrices, which is sharper than the conventional matrix Bernstein inequality and may be of independent interest.

math.ST↗

arXiv2014

A Generic Sample Splitting Approach for Refined Community Recovery in Stochastic Block Models

We propose and analyze a generic method for community recovery in stochastic block models and degree corrected block models. This approach can exactly recover the hidden communities with high probability when the expected node degrees are of order $\log n$ or higher. Starting from a roughly correct community partition given by some conventional community recovery algorithm, this method refines the partition in a cross clustering step. Our results simplify and extend some of the previous work on exact community recovery, discovering the key role played by sample splitting. The proposed method is simple and can be implemented with many practical community recovery algorithms.

stat.ML↗

arXiv2014

The functional mean-shift algorithm for mode hunting and clustering in infinite dimensions

We introduce the functional mean-shift algorithm, an iterative algorithm for estimating the local modes of a surrogate density from functional data. We show that the algorithm can be used for cluster analysis of functional data. We propose a test based on the bootstrap for the significance of the estimated local modes of the surrogate density. We present two applications of our methodology. In the first application, we demonstrate how the functional mean-shift algorithm can be used to perform spike sorting, i.e. cluster neural activity curves. In the second application, we use the functional mean-shift algorithm to distinguish between original and fake signatures.

stat.ME↗

arXiv2014

Functional Regression for Quasar Spectra

The Lyman-alpha forest is a portion of the observed light spectrum of distant galactic nuclei which allows us to probe remote regions of the Universe that are otherwise inaccessible. The observed Lyman-alpha forest of a quasar light spectrum can be modeled as a noisy realization of a smooth curve that is affected by a `damping effect' which occurs whenever the light emitted by the quasar travels through regions of the Universe with higher matter concentration. To decode the information conveyed by the Lyman-alpha forest about the matter distribution, we must be able to separate the smooth `continuum' from the noise and the contribution of the damping effect in the quasar light spectra. To predict the continuum in the Lyman-alpha forest, we use a nonparametric functional regression model in which both the response and the predictor variable (the smooth part of the damping-free portion of the spectrum) are function-valued random variables. We demonstrate that the proposed method accurately predicts the unobservable continuum in the Lyman-alpha forest both on simulated spectra and real spectra. Also, we introduce distribution-free prediction bands for the nonparametric functional regression model that have finite sample guarantees. These prediction bands, together with bootstrap-based confidence bands for the projection of the mean continuum on a fixed number of principal components, allow us to assess the degree of uncertainty in the model predictions.

stat.ME↗

arXiv2014

Minimax sparse principal subspace estimation in high dimensions

We study sparse principal components analysis in high dimensions, where $p$ (the number of variables) can be much larger than $n$ (the number of observations), and analyze the problem of estimating the subspace spanned by the principal eigenvectors of the population covariance matrix. We introduce two complementary notions of $\ell_q$ subspace sparsity: row sparsity and column sparsity. We prove nonasymptotic lower and upper bounds on the minimax subspace estimation error for $0\leq q\leq1$. The bounds are optimal for row sparse subspaces and nearly optimal for column sparse subspaces, they apply to general classes of covariance matrices, and they show that $\ell_q$ constrained estimates can achieve optimal minimax rates without restrictive spiked covariance conditions. Interestingly, the form of the rates matches known results for sparse regression when the effective noise variance is defined appropriately. Our proof employs a novel variational $\sinΘ$ theorem that may be useful in other regularized spectral estimation problems.

math.ST↗

arXiv2013

Squared-Norm Empirical Process in Banach Space

This note extends a recent result of Mendelson on the supremum of a quadratic process to squared norms of functions taking values in a Banach space. Our method of proof is a reduction by a symmetrization argument and observation about the subadditivity of the generic chaining functional. We provide an application to the supremum of a linear process in the sample covariance matrix indexed by finite rank, positive definite matrices.

math.PR↗

arXiv2013

A Conformal Prediction Approach to Explore Functional Data

This paper applies conformal prediction techniques to compute simultaneous prediction bands and clustering trees for functional data. These tools can be used to detect outliers and clusters. Both our prediction bands and clustering trees provide prediction sets for the underlying stochastic process with a guaranteed finite sample behavior, under no distributional assumptions. The prediction sets are also informative in that they correspond to the high density region of the underlying process. While ordinary conformal prediction has high computational cost for functional data, we use the inductive conformal predictor, together with several novel choices of conformity scores, to simplify the computation. Our methods are illustrated on some real data examples.

stat.ML↗

arXiv2012

Distribution Free Prediction Bands

We study distribution free, nonparametric prediction bands with a special focus on their finite sample behavior. First we investigate and develop different notions of finite sample coverage guarantees. Then we give a new prediction band estimator by combining the idea of "conformal prediction" (Vovk et al. 2009) with nonparametric conditional density estimation. The proposed estimator, called COPS (Conformal Optimized Prediction Set), always has finite sample guarantee in a stronger sense than the original conformal prediction estimator. Under regularity conditions the estimator converges to an oracle band at a minimax optimal rate. A fast approximation algorithm and a data driven method for selecting the bandwidth are developed. The method is illustrated first in simulated data. Then, an application shows that the proposed method gives desirable prediction intervals in an automatic way, as compared to the classical linear regression modeling.

stat.ME↗

arXiv2012

Minimax Rates of Estimation for Sparse PCA in High Dimensions

We study sparse principal components analysis in the high-dimensional setting, where $p$ (the number of variables) can be much larger than $n$ (the number of observations). We prove optimal, non-asymptotic lower and upper bounds on the minimax estimation error for the leading eigenvector when it belongs to an $\ell_q$ ball for $q \in [0,1]$. Our bounds are sharp in $p$ and $n$ for all $q \in [0, 1]$ over a wide class of distributions. The upper bound is obtained by analyzing the performance of $\ell_q$-constrained PCA. In particular, our results provide convergence rates for $\ell_1$-constrained PCA.

stat.ML↗

arXiv2011

Efficient Nonparametric Conformal Prediction Regions

We investigate and extend the conformal prediction method due to Vovk,Gammerman and Shafer (2005) to construct nonparametric prediction regions. These regions have guaranteed distribution free, finite sample coverage, without any assumptions on the distribution or the bandwidth. Explicit convergence rates of the loss function are established for such regions under standard regularity conditions. Approximations for simplifying implementation and data driven bandwidth selection methods are also discussed. The theoretical properties of our method are demonstrated through simulations.

math.ST↗