Recovering a low rank matrix from a subset of its entries, some of which may be corrupted, is known as the robust matrix completion (RMC) problem. Existing RMC methods have several limitations: they require a relatively large number of observed entries; they may fail under overparametrization, when their assumed rank is higher than the correct one; and many of them fail to recover even mildly ill-conditioned matrices. In this paper we propose a novel RMC method, denoted $\texttt{RGNMR}$, which overcomes these limitations. $\texttt{RGNMR}$ is a simple factorization-based iterative algorithm, which combines a Gauss–Newton linearization with removal of entries suspected to be outliers. On the theoretical front, we prove that under suitable assumptions, $\texttt{RGNMR}$ is guaranteed exact recovery of the underlying low rank matrix. Our theoretical results improve upon the best currently known for factorization-based methods. On the empirical front, we show via several simulations the advantages of $\texttt{RGNMR}$ over existing RMC methods, and in particular its ability to handle a small number of observed entries, overparameterization of the rank and ill-conditioned matrices. In addition, we propose a novel scheme for estimating the number of corrupted entries. This scheme may be used by other RMC methods that require as input the number of corrupted entries.

We study distributed schemes for high-dimensional sparse linear regression, based on orthogonal matching pursuit (OMP). Such schemes are particularly suited for settings where a central fusion center is connected to end machines, that have both computation and communication limitations. We prove that under suitable assumptions, distributed-OMP schemes recover the support of the regression vector with communication per machine linear in its sparsity and logarithmic in the dimension. Remarkably, this holds even at low signal-to-noise-ratios, where individual machines are unable to detect the support. Our simulations show that distributed-OMP schemes are competitive with more computationally intensive methods, and in some cases even outperform them.

The premise of semi-supervised learning (SSL) is that combining labeled and unlabeled data yields significantly more accurate models. Despite empirical successes, the theoretical understanding of SSL is still far from complete. In this work, we study SSL for high dimensional sparse Gaussian classification. To construct an accurate classifier a key task is feature selection, detecting the few variables that separate the two classes. For this SSL setting, we analyze information theoretic lower bounds for accurate feature selection as well as computational lower bounds, assuming the low-degree likelihood hardness conjecture. Our key contribution is the identification of a regime in the problem parameters (dimension, sparsity, number of labeled and unlabeled samples) where SSL is guaranteed to be advantageous for classification. Specifically, there is a regime where it is possible to construct in polynomial time an accurate SSL classifier. However, any computationally efficient supervised or unsupervised learning schemes, that separately use only the labeled or unlabeled data would fail. Our work highlights the provable benefits of combining labeled and unlabeled data for classification and feature selection in high dimensions. We present simulations that complement our theoretical analysis.

Large Language Models (LLMs) have revolutionised natural language processing, exhibiting impressive human-like capabilities. In particular, LLMs are capable of "lying", knowingly outputting false statements. Hence, it is of interest and importance to develop methods to detect when LLMs lie. Indeed, several authors trained classifiers to detect LLM lies based on their internal model activations. However, other researchers showed that these classifiers may fail to generalise, for example to negated statements. In this work, we aim to develop a robust method to detect when an LLM is lying. To this end, we make the following key contributions: (i) We demonstrate the existence of a two-dimensional subspace, along which the activation vectors of true and false statements can be separated. Notably, this finding is universal and holds for various LLMs, including Gemma-7B, LLaMA2-13B, Mistral-7B and LLaMA3-8B. Our analysis explains the generalisation failures observed in previous studies and sets the stage for more robust lie detection; (ii) Building upon (i), we construct an accurate LLM lie detector. Empirically, our proposed classifier achieves state-of-the-art performance, attaining 94\% accuracy in both distinguishing true from false factual statements and detecting lies generated in real-world scenarios.

We consider the problem of mixed linear regression (MLR), where each observed sample belongs to one of $K$ unknown linear models. In practical applications, the mixture of the $K$ models may be imbalanced with a significantly different number of samples from each model. Unfortunately, most MLR methods do not perform well in such settings. Motivated by this practical challenge, in this work we propose Mix-IRLS, a novel, simple and fast algorithm for MLR with excellent performance on both balanced and imbalanced mixtures. In contrast to popular approaches that recover the $K$ models simultaneously, Mix-IRLS does it sequentially using tools from robust regression. Empirically, beyond imbalanced mixtures, Mix-IRLS succeeds in a broad range of additional settings where other methods fail, including small sample sizes, presence of outliers, and an unknown number of models $K$. Furthermore, Mix-IRLS outperforms competing methods on several real-world datasets, in some cases by a large margin. We complement our empirical results by deriving a recovery guarantee for Mix-IRLS, which highlights its advantage on imbalanced mixtures.

Merging the predictions of multiple experts is a frequent task. When ground-truth response values are available, this merging is often based on the estimated accuracies of the experts. In various applications, however, the only available information are the experts’ predictions on unlabeled test data, which do not allow to directly estimate their accuracies. Moreover, simple merging schemes such as majority voting in classification or the ensemble mean or median in regression, are clearly sub-optimal when some experts are more accurate than others. Focusing on regression tasks, in this work we propose U-PCR, a framework for unsupervised ensemble regression. Specifically, we develop spectral-based methods that under mild assumptions and in the absence of ground truth data, are able to estimate the mean squared error of the different experts and combine their predictions to a more accurate meta-learner. We provide theoretical support for U-PCR as well as empirical evidence for the validity of its underlying assumptions. On a variety of regression problems, we illustrate the improved accuracy of U-PCR over various unsupervised merging strategies. Finally, we also illustrate its applicability to unsupervised multi-class ensemble learning.

The inductive matrix completion (IMC) problem is to recover a low rank matrix from few observed entries while incorporating prior knowledge about its row and column subspaces. In this work, we make three contributions to the IMC problem: (i) we prove that under suitable conditions, the IMC optimization landscape has no bad local minima; (ii) we derive a simple scheme with theoretical guarantees to estimate the rank of the unknown matrix; and (iii) we propose GNIMC, a simple Gauss-Newton based method to solve the IMC problem, analyze its runtime and derive for it strong recovery guarantees. The guarantees for GNIMC are sharper in several aspects than those available for other methods, including a quadratic convergence rate, fewer required observed entries and stability to errors or deviations from low-rank. Empirically, given entries observed uniformly at random, GNIMC recovers the underlying matrix substantially faster than several competing methods.

We study the blind calibration problem of uniform linear arrays of acoustic vector sensors for narrowband Gaussian signals, and propose an improved, asymptotically optimal blind calibration scheme. Following recent work by Ramamohan et al., we exploit the special (block-Toeplitz) structure of the underlying signals' spatial covariance matrix. However, we offer a substantial improvement over their ordinary Least Squares (LS)-based approach: Using asymptotic approximations we obtain Optimally-Weighted LS estimates of the sensors' gains and phases offsets. We show via simulations that our estimates exhibit near-optimal performance, with improvements reaching more than an order of magnitude in the mean squared estimation errors of the calibration parameters, as well as in directions of-arrival estimation.

Several classification methods assume that the underlying distributions follow tree-structured graphical models. Indeed, trees capture statistical dependencies between pairs of variables, which may be crucial to attaining low classification errors. In this setting, the optimal classifier is linear in the log-transformed univariate and bivariate densities that correspond to the tree edges. In practice, observed data may not be well approximated by trees. Yet, motivated by the importance of pairwise dependencies for accurate classification, here we propose to approximate the optimal decision boundary by a sparse linear combination of the univariate and bivariate log-transformed densities. Our proposed approach is semi-parametric in nature: we non-parametrically estimate the univariate and bivariate densities, remove pairs of variables that are nearly independent using the Hilbert-Schmidt independence criterion, and finally construct a linear SVM using the retained log-transformed densities. We demonstrate on synthetic and real data sets, that our classifier, named SLB (sparse log-bivariate density), is competitive with other popular classification methods.

A fundamental question for edge detection in noisy images is how faint can an edge be and still be detected. In this paper we offer a formalism to study this question and subsequently introduce computationally efficient multiscale edge detection algorithms designed to detect faint edges in noisy images. In our formalism we view edge detection as a search in a discrete, though potentially large, set of feasible curves. First, we derive approximate expressions for the detection threshold as a function of curve length and the complexity of the search space. We then present two edge detection algorithms, one for straight edges, and the second for curved ones. Both algorithms efficiently search for edges in a large set of candidates by hierarchically constructing difference filters that match the curves traced by the sought edges. We demonstrate the utility of our algorithms in both simulations and applications involving challenging real images. Finally, based on these principles, we develop an algorithm for fiber detection and enhancement. We exemplify its utility to reveal and enhance nerve axons in light microscopy images.

Latent variable models with hidden binary units appear in various applications. Learning such models, in particular in the presence of noise, is a challenging computational problem. In this paper we propose a novel spectral approach to this problem, based on the eigenvectors of both the second order moment matrix and third order moment tensor of the observed data. We prove that under mild non-degeneracy conditions, our method consistently estimates the model parameters at the optimal parametric rate. Our tensor-based method generalizes previous orthogonal tensor decomposition approaches, where the hidden units were assumed to be either statistically independent or mutually exclusive. We illustrate the consistency of our method on simulated data and demonstrate its usefulness in learning a common model for population mixtures in genetics.

We consider semi-supervised regression when the predictor variables are drawn from an unknown manifold. A simple two step approach to this problem is to: (i) estimate the manifold geodesic distance between any pair of points using both the labeled and unlabeled instances; and (ii) apply a k nearest neighbor regressor based on these distance estimates. We prove that given sufficiently many unlabeled points, this simple method of geodesic kNN regression achieves the optimal finite-sample minimax bound on the mean squared error, as if the manifold were known. Furthermore, we show how this approach can be efficiently implemented, requiring only O(k N log N) operations to estimate the regression function at all N labeled and unlabeled points. We illustrate this approach on two datasets with a manifold structure: indoor localization using WiFi fingerprints and facial pose estimation. In both cases, geodesic kNN is more accurate and much faster than the popular Laplacian eigenvector regressor.

Many applications require learning classifiers or regressors that are both accurate and cheap to evaluate. Prediction cost can be drastically reduced if the learned predictor is constructed such that on the majority of the inputs, it uses cheap features and fast evaluations. The main challenge is to do so with little loss in accuracy. In this work we propose a budget-aware strategy based on deep boosted regression trees. In contrast to previous approaches to learning with cost penalties, our method can grow very deep trees that on average are nonetheless cheap to compute. We evaluate our method on a number of datasets and find that it outperforms the current state of the art by a large margin. Our algorithm is easy to implement and its learning time is comparable to that of the original gradient boosting. Source code is made available at http://github.com/svenpeter42/LightGBM-CEGB.

In unsupervised ensemble learning, one obtains predictions from multiple sources or classifiers, yet without knowing the reliability and expertise of each source, and with no labeled data to assess it. The task is to combine these possibly conflicting predictions into an accurate meta-learner. Most works to date assumed perfect diversity between the different sources, a property known as conditional independence. In realistic scenarios, however, this assumption is often violated, and ensemble learners based on it can be severely sub-optimal. The key challenges we address in this paper are: (i) how to detect, in an unsupervised manner, strong violations of conditional independence; and (ii) construct a suitable meta-learner. To this end we introduce a statistical model that allows for dependencies between classifiers. Based on this model, we develop novel unsupervised methods to detect strongly dependent classifiers, better estimate their accuracies, and construct an improved meta-learner. Using both artificial and real datasets, we showcase the importance of taking classifier dependencies into account and the competitive performance of our approach.

Detecting edges is a fundamental problem in computer vision with many applications, some involving very noisy images. While most edge detection methods are fast, they perform well only on relatively clean images. Unfortunately, sophisticated methods that are robust to high levels of noise are quite slow. In this paper we develop a novel multiscale method to detect curved edges in noisy images. Even though our algorithm searches for edges over an exponentially large set of candidate curves, its runtime is nearly linear in the total number of image pixels. As we demonstrate experimentally, our algorithm is orders of magnitude faster than previous methods designed to deal with high noise levels. At the same time it obtains comparable and often superior results to existing methods on a variety of challenging noisy images.

We show how deep learning methods can be applied in the context of crowdsourcing and unsupervised ensemble learning. First, we prove that the popular model of Dawid and Skene, which assumes that all classifiers are conditionally independent, is \em equivalent to a Restricted Boltzmann Machine (RBM) with a single hidden node. Hence, under this model, the posterior probabilities of the true labels can be instead estimated via a trained RBM. Next, to address the more general case, where classifiers may strongly violate the conditional independence assumption, we propose to apply RBM-based Deep Neural Net (DNN). Experimental results on various simulated and real-world datasets demonstrate that our proposed DNN approach outperforms other state-of-the-art methods, in particular when the data violates the conditional independence assumption.

In various situations one is given only the predictions of multiple classifiers over a large unlabeled test data. This scenario raises the following questions: Without any labeled data and without any a-priori knowledge about the reliability of these different classifiers, is it possible to consistently and computationally efficiently estimate their accuracies? Furthermore, also in a completely unsupervised manner, can one construct a more accurate unsupervised ensemble classifier? In this paper, focusing on the binary case, we present simple, computationally efficient algorithms to solve these questions. Furthermore, under standard classifier independence assumptions, we prove our methods are consistent and study their asymptotic error. Our approach is spectral, based on the fact that the off-diagonal entries of the classifiers’ covariance matrix and 3-d tensor are rank-one. We illustrate the competitive performance of our algorithms via extensive experiments on both artificial and real datasets.

In various applications involving hidden Markov models (HMMs), some of the hidden states are aliased, having identical output distributions. The minimality, identifiability and learnability of such aliased HMMs have been long standing problems, with only partial solutions provided thus far. In this paper we focus on parametric-output HMMs, whose output distributions come from a parametric family, and that have exactly two aliased states. For this class, we present a complete characterization of their minimality and identifiability. Furthermore, for a large family of parametric output distributions, we derive computationally efficient and statistically consistent algorithms to detect the presence of aliasing and learn the aliased HMM transition and emission parameters. We illustrate our theoretical analysis by several simulations.

Edge detection is a challenging, important task in image analysis. Various applications require real-time detection of long edges in large and noisy images, possibly under limited computational resources. While standard edge detection methods are computationally fast, they perform well only at low levels of noise. Modern sophisticated methods, in contrast, are robust to noise, but may be too slow for real-time processing of large images. This raises the following question, which is the focus of our paper: How well can one detect long edges in noisy images under severe computational constraints that allow only a fraction of all image pixels to be processed? We make several theoretical and practical contributions regarding this problem. We develop possibly the first sublinear algorithm to detect long straight edges in noisy images. In addition, we theoretically analyze the inevitable tradeoff between its detection performance and the allowed computational budget. Finally, we demonstrate its competitive performance on both simulated and real images.

Over the past decade, single image Super-Resolution (SR) research has focused on developing sophisticated image priors, leading to significant advances. Estimating and incorporating the blur model, that relates the high-res and low-res images, has received much less attention, however. In particular, the reconstruction constraint, namely that the blurred and down sampled high-res output should approximately equal the low-res input image, has been either ignored or applied with default fixed blur models. In this work, we examine the relative importance of the image prior and the reconstruction constraint. First, we show that an accurate reconstruction constraint combined with a simple gradient regularization achieves SR results almost as good as those of state-of-the-art algorithms with sophisticated image priors. Second, we study both empirically and theoretically the sensitivity of SR algorithms to the blur model assumed in the reconstruction constraint. We find that an accurate blur model is more important than a sophisticated image prior. Finally, using real camera data, we demonstrate that the default blur models of various SR algorithms may differ from the camera blur, typically leading to over-smoothed results. Our findings highlight the importance of accurately estimating camera blur in reconstructing raw lowers images acquired by an actual camera.

We present a novel approach to learning an HMM whose outputs are distributed according to a parametric family. This is done by \em decoupling the learning task into two steps: first estimating the output parameters, and then estimating the hidden states transition probabilities. The first step is accomplished by fitting a mixture model to the output stationary distribution. Given the parameters of this mixture model, the second step is formulated as the solution of an easily solvable convex quadratic program. We provide an error analysis for the estimated transition probabilities and show they are robust to small perturbations in the estimates of the mixture parameters. Finally, we support our analysis with some encouraging empirical results.

The goal of natural image denoising is to estimate a clean version of a given noisy image, utilizing prior knowledge on the statistics of natural images. The problem has been studied intensively with considerable progress made in recent years. However, it seems that image denoising algorithms are starting to converge and recent algorithms improve over previous ones by only fractional dB values. It is thus important to understand how much more can we still improve natural image denoising algorithms and what are the inherent limits imposed by the actual statistics of the data. The challenge in evaluating such limits is that constructing proper models of natural image statistics is a long standing and yet unsolved problem. To overcome the absence of accurate image priors, this paper takes a non parametric approach and represents the distribution of natural images using a huge set of 1010patches. We then derive a simple statistical measure which provides a lower bound on the optimal Bayesian minimum mean square error (MMSE). This imposes a limit on the best possible results of denoising algorithms which utilize a fixed support around a denoised pixel and a generic natural image prior. Our findings suggest that for small windows, state of the art denoising algorithms are approaching optimality and cannot be further improved beyond ~ 0.1dB values.

In this paper we consider signal detection in cognitive radio networks, under a non-parametric, multi-sensor detection scenario, and compare the cases of known and unknown noise level. The analysis is focused on two eigenvalue-based methods, namely Roy's largest root test, which requires knowledge of the noise variance, and the generalized likelihood ratio test, which can be interpreted as a test of the largest eigenvalue vs. a maximum-likelihood estimate of the noise variance. The detection performance of the two considered methods is expressed by closed-form analytical formulas, shown to be accurate even for small number of sensors and samples. We then derive an expression of the gap between the two detectors in terms of the signal-to-noise ratio of the signal to be detected, and we identify critical settings where this gap is significant (e.g., low number of sensors and signal strength). Our results thus provide a measure of the impact of noise level knowledge and highlight the importance of accurate noise estimation.

We study the behavior of the popular Laplacian Regularization method for Semi-Supervised Learning at the regime of a fixed number of labeled points but a large number of unlabeled points. We show that in $\R^d$, $d \geq 2$, the method is actually not well-posed, and as the number of unlabeled points increases the solution degenerates to a noninformative function. We also contrast the method with the Laplacian Eigenvector method, and discuss the ``smoothness assumptions associated with this alternate method.

Nonlocal neighborhood filters are modern and powerful techniques for image and signal denoising. In this paper, we give a probabilistic interpretation and analysis of the method viewed as a random walk on the patch space. We show that the method is intimately connected to the characteristics of diffusion processes, their escape times over potential barriers, and their spectral decomposition. In particular, the eigenstructure of the diffusion operator leads to novel insights on the performance and limitations of the denoising method, as well as a proposal for an improved filtering algorithm.

Two-tone (ldquoMooneyrdquo) images seem to arouse vivid 3D percept of faces, both familiar and unfamiliar, despite their seemingly poor content. Recent psychological and fMRI studies suggest that this percept is guided primarily by top-down procedures in which recognition precedes reconstruction. In this paper we investigate this hypothesis from a mathematical standpoint. We show that indeed, under standard shape from shading assumptions, a Mooney image can give rise to multiple different 3D reconstructions even if reconstruction is restricted to the Mooney transition curve (the boundary curve between black and white) alone. We then use top-down reconstruction methods to recover the shape of novel faces from single Mooney images exploiting prior knowledge of the structure of at least one face of a different individual. We apply these methods to thresholded images of real faces and compare the reconstruction quality relative to reconstruction from gray level images.

Spectral clustering methods are common graph-based approaches to clustering of data. Spectral clustering algorithms typically start from local information encoded in a weighted graph on the data and cluster according to the global eigenvectors of the corresponding (normalized) similarity matrix. One contribution of this paper is to present fundamental limitations of this general local to global approach. We show that based only on local information, the normalized cut functional is not a suitable measure for the quality of clustering. Further, even with a suitable similarity measure, we show that the first few eigenvectors of such adjacency matrices cannot successfully cluster datasets that contain structures at different scales of size and density. Based on these findings, a second contribution of this paper is a novel diffusion based measure to evaluate the coherence of individual clusters. Our measure can be used in conjunction with any bottom-up graph-based clustering method, it is scale-free and can determine coherent clusters at all scales. We present both synthetic examples and real image segmentation problems where various spectral clustering algorithms fail. In contrast, using this coherence measure finds the expected clusters at all scales. Keywords: Clustering, kernels, learning theory.

This paper presents a diffusion based probabilistic interpretation of spectral clustering and dimensionality reduction algorithms that use the eigenvectors of the normalized graph Laplacian. Given the pairwise adjacency matrix of all points, we define a diffusion distance between any two data points and show that the low dimensional representation of the data by the first few eigenvectors of the corresponding Markov matrix is optimal under a certain mean squared error criterion. Furthermore, assuming that data points are random samples from a density p(x) = e-U (x) we identify these eigenvectors as discrete approximations of eigenfunctions of a Fokker-Planck operator in a potential 2U (x) with reflecting boundary conditions. Finally, applying known results regarding the eigenvalues and eigenfunctions of the continuous Fokker-Planck operator, we provide a mathematical justification for the success of spectral clustering and dimensional reduction algorithms based on these first few eigenvectors. This analysis elucidates, in terms of the characteristics of diffusion processes, many empirical findings regarding spectral clustering algorithms. Keywords: Algorithms and architectures, learning theory.