Uri Shaham

Yale University | YU · Department of Statistics

Modern datasets often contain large subsets of correlated features and nuisance features, which are not or loosely related to the main underlying structures of the data. Nuisance features can be identified using the Laplacian score criterion, which evaluates the importance of a given feature via its consistency with the Graph Laplacians’ leading ei...

Latent variable discovery is a central problem in data analysis with a broad range of applications in applied science. In this work, we consider data given as an invertible mixture of two statistically independent components, and assume that one of the components is observed while the other is hidden. Our goal is to recover the hidden component. Fo...

Modern datasets often contain large subsets of correlated features and nuisance features, which are not or loosely related to the main underlying structures of the data. Nuisance features can be identified using the Laplacian score criterion, which evaluates the importance of a given feature via its consistency with the Graph Laplacians' leading ei...

We propose a framework for deep ordinal regression, based on unimodal output distribution and optimal transport loss. Despite being seemingly appropriate, in many recent works the unimodality requirement is either absent, or implemented using soft targets, which do not guarantee unimodal outputs at inference. In addition, we argue that the standard...

Scientific observations often consist of a large number of variables (features). Identifying a subset of meaningful features is often ignored in unsupervised learning, despite its potential for unraveling clear patterns hidden in the ambient space. In this paper, we present a method for unsupervised feature selection, tailored for the task of clust...

Biological measurements often contain systematic errors, also known as "batch effects", which may invalidate downstream analysis when not handled correctly. The problem of removing batch effects is of major importance in the biological community. Despite recent advances in this direction via deep learning techniques, most current methods may not fu...

We study the effectiveness of various approaches that defend against adversarial attacks on deep networks via manipulations based on basis function representations of images. Specifically, we experiment with low-pass filtering, PCA, JPEG compression, low resolution wavelet approximation, and soft-thresholding. We evaluate these defense techniques u...

Background: Medical practitioners use survival models to explore and understand the relationships between patients' covariates (e.g. clinical and genetic features) and the effectiveness of various treatment options. Standard survival models like the linear Cox proportional hazards model require extensive feature engineering or prior medical knowle...

Spectral clustering is a leading and popular technique in unsupervised data analysis. Two of its major limitations are scalability and generalization of the spectral embedding (i.e., out-of-sample-extension). In this paper we introduce a deep learning approach to spectral clustering that overcomes the above shortcomings. Our network, which we call...

We consider the statistical problem of learning a common source of variability in data which are synchronously captured by multiple sensors, and demonstrate that Siamese neural networks can be naturally applied to this problem. This approach is useful in particular in exploratory, data-driven applications, where neither a model nor label informatio...

Motivation: Mass cytometry or CyTOF is an emerging technology for high-dimensional multiparameter single cell analysis that overcomes many limitations of fluorescence-based flow cytometry. New methods for analyzing CyTOF data attempt to improve automation, scalability, performance and interpretation of data generated in large studies. Assigning ind...

Stochastic Neighbor Embedding and its variants are widely used dimensionality reduction techniques -- despite their popularity, no theoretical results are known. We prove that the optimal SNE embedding of well-separated clusters from high dimensions to the real line $\mathbb{R}$ manages to successfully separate the clusters in a quantitative way.

Tumor genomes evolve through a selection of mutations. These mutations may complement each other to promote tumorigenesis. To better understand the functional interactions of different processes in cancer, we studied mutation data of a set of tumors and identified significantly co-mutated pathways. Fisher’s exact test is a standard approach that ca...

Motivation: Sources of variability in experimentally derived data include measurement error in addition to the physical phenomena of interest. This measurement error is a combination of systematic components, originating from the measuring instrument and random measurement errors. Several novel biological technologies, such as mass cytometry and s...

Previous research has shown that neural networks can model survival data in situations in which some patients' death times are unknown, e.g. right-censored. However, neural networks have rarely been shown to outperform their linear counterparts such as the Cox proportional hazards model. In this paper, we run simulated experiments and use real surv...

Mass cytometry or CyTOF is an emerging technology for high-dimensional multiparameter single cell analysis that overcomes many limitations of fluorescence-based flow cytometry. New methods for analyzing CyTOF data attempt to improve automation, scalability, performance, and interpretation of data generated in large studies. Assigning individual cel...

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 mod...

We propose an Artificial Neural Network (ANN) approach for discovering common hidden variables and for learning of invariant representations. This approach uses synchronicity and concurrence to recognize desired structures in data and discard superfluous structures, in the absence of labels and a data generation model. In the common variable discov...

We propose a general framework for increasing local stability of Artificial Neural Nets (ANNs) using Robust Optimization (RO). We achieve this through an alternating minimization-maximization procedure, in which the loss of the network is minimized over perturbed examples that are generated at each parameter update. We show that adversarial trainin...

We discuss approximation of functions using deep neural nets. Given a function $f$ on a $d$-dimensional manifold $\Gamma \subset \mathbb{R}^m$, we construct a sparsely-connected depth-4 neural network and bound its error in approximating $f$. The size of the network depends on dimension and curvature of the manifold $\Gamma$, the complexity of $f$,...

Non-linear manifold learning enables high-dimensional data analysis, but requires out-of-sample-extension methods to process new data points. In this paper, we propose a manifold learning algorithm based on deep learning to create an encoder, which maps a high-dimensional dataset and its low-dimensional embedding, and a decoder, which takes the emb...

Pain has been increasingly studied as for its importance in patient care and recovery. In the past, several ECG (electrocardiogram) signal processing methods have been studied for the detection and classification of pain without achieving clinical application. In this work we study the ECG signals from healthy adults during acute pain, and perform...

Classification algorithms that use the NCD (Normalized Compression Distance) as a similarity metric are proposed. This way of measuring similarity allows either skipping the feature selection and feature extraction phases or extracting features in a more objective way than common feature extraction methods, and so makes the classification algorithm...