Source author record

Hariharan Narayanan

Explore this author’s publications, collaborators and research topics. This catalog record has not been claimed by its author.

ResearcherUnclaimed source record

Machine Learning Data Structures and Algorithms math.PR Computational Geometry Computation math.CO math.ST Statistics Theory eess.SY math.CA math.DG math.OC Methodology Numerical Analysis Systems and Control

Catalog footprint

Research overview

15works

15topics

4close collaborators

Actions

Is this your work?

Open graph Browse papers

preprint2026arXiv

Denoising data using convex relaxations

We study the problem of denoising observations $Y_i=X_i+Z_i$ , where the latent variables $X_i$ are sampled from a low-dimensional manifold in $\mathbb{R}^n$ and the noise variables $Z_i$ are isotropic Gaussian. We propose a convex-relaxation estimator that first reduces dimension by principal component analysis and then projects the observations onto the convex hull of the projected latent manifold. We construct a statistical oracle that estimates its supporting hyperplanes from empirical Gaussian tail probabilities of the noisy sample. Under a lower-mass condition on the latent distribution, we prove finite-sample guarantees for the oracle and derive error bounds for the resulting denoiser. The analysis combines risk bounds for least-squares projection under convex constraints with entropy bounds for convex hulls. We also verify the assumptions of the framework for a Cryo-Electron Microscopy observation model by establishing suitable covering number and Lipschitz estimates for the associated group action and imaging operators.

preprint2022arXiv

Fitting a manifold of large reach to noisy data

Let ${\mathcal M}\subset {\mathbb R}^n$ be a $C^2$ -smooth compact submanifold of dimension $d$ . Assume that the volume of ${\mathcal M}$ is at most $V$ and the reach (i.e. the normal injectivity radius) of ${\mathcal M}$ is greater than $τ$ . Moreover, let $μ$ be a probability measure on ${\mathcal M}$ whose density on ${\mathcal M}$ is a strictly positive Lipschitz-smooth function. Let $x_j\in {\mathcal M}$ , $j=1,2,\dots,N$ be $N$ independent random samples from distribution $μ$ . Also, let $ξ_j$ , $j=1,2,\dots, N$ be independent random samples from a Gaussian random variable in ${\mathbb R}^n$ having covariance $σ^2I$ , where $σ$ is less than a certain specified function of $d, V$ and $τ$ . We assume that we are given the data points $y_j=x_j+ξ_j,$ $j=1,2,\dots,N$ , modelling random points of ${\mathcal M}$ with measurement noise. We develop an algorithm which produces from these data, with high probability, a $d$ dimensional submanifold ${\mathcal M}_o\subset {\mathbb R}^n$ whose Hausdorff distance to ${\mathcal M}$ is less than $Cdσ^2/τ$ and whose reach is greater than $cτ/d^6$ with universal constants $C,c > 0$ . The number $N$ of random samples required depends almost linearly on $n$ , polynomially on $σ^{-1}$ and exponentially on $d$ .

preprint2022arXiv

On the mixing time of coordinate Hit-and-Run

We obtain a polynomial upper bound on the mixing time $T_{CHR}(ε)$ of the coordinate Hit-and-Run random walk on an $n-$ dimensional convex body, where $T_{CHR}(ε)$ is the number of steps needed in order to reach within $ε$ of the uniform distribution with respect to the total variation distance, starting from a warm start (i.e., a distribution which has a density with respect to the uniform distribution on the convex body that is bounded above by a constant). Our upper bound is polynomial in $n, R$ and $\frac{1}ε$ , where we assume that the convex body contains the unit $\Vert\cdot\Vert_\infty$ -unit ball $B_\infty$ and is contained in its $R$ -dilation $R\cdot B_\infty$ . Whether coordinate Hit-and-Run has a polynomial mixing time has been an open question.

preprint2020arXiv

Implicit Linear Algebra and Basic Circuit Theory

In this paper we derive some basic results of circuit theory using `Implicit Linear Algebra' (ILA). This approach has the advantage of simplicity and generality. Implicit linear algebra is outlined in [1]. We denote the space of all vectors on $S$ by $\mathcal{F}_S$ and the space containing only the zero vector on $S$ by $\mathbf{0}_S.$ The dual $\mathcal{V}_S^{\perp}$ of a vector space $\mathcal{V}_S$ is the collection of all vectors whose dot product with vectors in $\mathcal{V}_S$ is zero. The basic operation of ILA is a linking operation ('matched composition`) between vector spaces $\mathcal{V}_{SP},\mathcal{V}_{PQ}$ (regarded as collections of row vectors on column sets $S\cup P, P\cup Q,$ respectively with $S,P,Q$ disjoint) defined by $\mathcal{V}_{SP}\leftrightarrow \mathcal{V}_{PQ}\equiv \{(f_S,h_Q):((f_S,g_P)\in \mathcal{V}_{SP}, (g_P,h_Q) \in \mathcal{V}_{PQ}\},$ and another ('skewed composition`) defined by $\mathcal{V}_{SP}\rightleftharpoons \mathcal{V}_{PQ}\equiv \{(f_S,h_Q):((f_S,g_P)\in \mathcal{V}_{SP}, (-g_P,h_Q) \in \mathcal{V}_{PQ}\}.$ The basic results of ILA are the Implicit Inversion Theorem (which states that $\mathcal{V}_{SP}\leftrightarrow(\mathcal{V}_{SP}\leftrightarrow \mathcal{V}_S)= \mathcal{V}_S,$ iff $\mathcal{V}_{SP}\leftrightarrow \mathbf{0}_P\subseteq \mathcal{V}_S\subseteq \mathcal{V}_{SP}\leftrightarrow\mathcal{F}_S$ ) and Implicit Duality Theorem (which states that $(\mathcal{V}_{SP}\leftrightarrow \mathcal{V}_{PQ})^{\perp}= (\mathcal{V}_{SP}^{\perp}\rightleftharpoons \mathcal{V}_{PQ}^{\perp}$ ). We show that the operations and results of ILA are useful in understanding basic circuit theory. We illustrate this by using ILA to present a generalization of Thevenin-Norton theorem where we compute multiport behaviour using adjoint multiport termination through a gyrator and a very general version of maximum power transfer theorem, which states that the port conditions that appear, during adjoint multiport termination through an ideal transformer, correspond to maximum power transfer.

preprint2020arXiv

John's Walk

We present an affine-invariant random walk for drawing uniform random samples from a convex body $\mathcal{K} \subset \mathbb{R}^n$ that uses maximum volume inscribed ellipsoids, known as John's ellipsoids, for the proposal distribution. Our algorithm makes steps using uniform sampling from the John's ellipsoid of the symmetrization of $\mathcal{K}$ at the current point. We show that from a warm start, the random walk mixes in $\widetilde{O}(n^7)$ steps where the log factors depend only on constants associated with the warm start and desired total variation distance to uniformity. We also prove polynomial mixing bounds starting from any fixed point $x$ such that for any chord $pq$ of $\mathcal{K}$ containing $x$ , $\left|\log \frac{|p-x|}{|q-x|}\right|$ is bounded above by a polynomial in $n$ .

preprint2020arXiv

Random concave functions on an equilateral lattice with periodic Hessians I: entropy and Laplacians

We show that a random concave function having a periodic hessian on an equilateral lattice has a quadratic scaling limit, if the average hessian of the function satisfies certain conditions. We consider the set of all concave functions $g$ on an equilateral lattice $\mathbb L$ that when shifted by an element of $n \mathbb L$ , incur addition by a linear function (this condition is equivalent to the periodicity of the hessian of $g$ ). We identify this set, up to addition by a constant, with a convex polytope $P_n(s)$ , where $s$ corresponds to the average hessian. We show that the $\ell_\infty$ diameter of $P_n(s)$ is bounded below by $c(s) n^2$ , where $c(s)$ is a positive constant depending only on $s$ . Our main result is that, for any $ε_0 > 0$ , the normalized Lebesgue measure of all points in $P_n(s)$ that are not contained in a $n^2$ dimensional cube $Q$ of sidelength $2 ε_0 n^2$ , centered at the unique (up to addition of a linear term) quadratic polynomial with hessian $s$ , tends to $0$ as $n$ tends to $\infty$ .

preprint2015arXiv

Escaping the Local Minima via Simulated Annealing: Optimization of Approximately Convex Functions

We consider the problem of optimizing an approximately convex function over a bounded convex set in $\mathbb{R}^n$ using only function evaluations. The problem is reduced to sampling from an \emph{approximately} log-concave distribution using the Hit-and-Run method, which is shown to have the same $\mathcal{O}^*$ complexity as sampling from log-concave distributions. In addition to extend the analysis for log-concave distributions to approximate log-concave distributions, the implementation of the 1-dimensional sampler of the Hit-and-Run walk requires new methods and analysis. The algorithm then is based on simulated annealing which does not relies on first order conditions which makes it essentially immune to local minima. We then apply the method to different motivating problems. In the context of zeroth order stochastic convex optimization, the proposed method produces an $ε$ -minimizer after $\mathcal{O}^*(n^{7.5}ε^{-2})$ noisy function evaluations by inducing a $\mathcal{O}(ε/n)$ -approximately log concave distribution. We also consider in detail the case when the "amount of non-convexity" decays towards the optimum of the function. Other applications of the method discussed in this work include private computation of empirical risk minimizers, two-stage stochastic programming, and approximate dynamic programming for online learning.

preprint2015arXiv

Randomized Interior Point methods for Sampling and Optimization

We present a Markov chain (Dikin walk) for sampling from a convex body equipped with a self-concordant barrier, whose mixing time from a "central point" is strongly polynomial in the description of the convex set. The mixing time of this chain is invariant under affine transformations of the convex set, thus eliminating the need for first placing the body in an isotropic position. This recovers and extends previous results of from polytopes to more general convex sets. On every convex set of dimension $n$ , there exists a self-concordant barrier whose "complexity" is polynomially bounded. Consequently, a rapidly mixing Markov chain of the kind we describe can be defined on any convex set. We use these results to design an algorithm consisting of a single random walk for optimizing a linear function on a convex set. We show that this random walk reaches an approximately optimal point in polynomial time with high probability and that the corresponding objective values converge with probability 1 to the optimal objective value as the number of steps tends to infinity. One technical contribution is a family of lower bounds for the isoperimetric constants of (weighted) Riemannian manifolds on which, interior point methods perform a kind of steepest descent. Using results of Barthe \cite{barthe} and Bobkov and Houdré, on the isoperimetry of products of (weighted) Riemannian manifolds, we obtain sharper upper bounds on the mixing time of Dikin walk on products of convex sets than the bounds obtained from a direct application of the Localization Lemma, on which, since (Lovász and Simonovits), the analyses of all random walks on convex sets have relied.

preprint2014arXiv

On Zeroth-Order Stochastic Convex Optimization via Random Walks

We propose a method for zeroth order stochastic convex optimization that attains the suboptimality rate of $\tilde{\mathcal{O}}(n^{7}T^{-1/2})$ after $T$ queries for a convex bounded function $f:{\mathbb R}^n\to{\mathbb R}$ . The method is based on a random walk (the \emph{Ball Walk}) on the epigraph of the function. The randomized approach circumvents the problem of gradient estimation, and appears to be less sensitive to noisy function evaluations compared to noiseless zeroth order methods.

preprint2013arXiv

Efficient Sampling from Time-Varying Log-Concave Distributions

We propose a computationally efficient random walk on a convex body which rapidly mixes and closely tracks a time-varying log-concave distribution. We develop general theoretical guarantees on the required number of steps; this number can be calculated on the fly according to the distance from and the shape of the next distribution. We then illustrate the technique on several examples. Within the context of exponential families, the proposed method produces samples from a posterior distribution which is updated as data arrive in a streaming fashion. The sampling technique can be used to track time-varying truncated distributions, as well as to obtain samples from a changing mixture model, fitted in a streaming fashion to data. In the setting of linear optimization, the proposed method has oracle complexity with best known dependence on the dimension for certain geometries. In the context of online learning and repeated games, the algorithm is an efficient method for implementing no-regret mixture forecasting strategies. Remarkably, in some of these examples, only one step of the random walk is needed to track the next distribution.

preprint2013arXiv

Estimating Certain Non-Zero Littlewood-Richardson Coefficients

Littlewood Richardson coefficients are structure constants appearing in the representation theory of the general linear groups ( $GL_n$ ). The main results of this paper are: 1. A strongly polynomial randomized approximation scheme for certain Littlewood-Richardson coefficients. 2. A proof of approximate log-concavity of certain Littlewood-Richardson coefficients.

preprint2013arXiv

Testing the Manifold Hypothesis

The hypothesis that high dimensional data tend to lie in the vicinity of a low dimensional manifold is the basis of manifold learning. The goal of this paper is to develop an algorithm (with accompanying complexity guarantees) for fitting a manifold to an unknown probability distribution supported in a separable Hilbert space, only using i.i.d samples from that distribution. More precisely, our setting is the following. Suppose that data are drawn independently at random from a probability distribution $P$ supported on the unit ball of a separable Hilbert space $H$ . Let $G(d, V, τ)$ be the set of submanifolds of the unit ball of $H$ whose volume is at most $V$ and reach (which is the supremum of all $r$ such that any point at a distance less than $r$ has a unique nearest point on the manifold) is at least $τ$ . Let $L(M, P)$ denote mean-squared distance of a random point from the probability distribution $P$ to $M$ . We obtain an algorithm that tests the manifold hypothesis in the following sense. The algorithm takes i.i.d random samples from $P$ as input, and determines which of the following two is true (at least one must be): (a) There exists $M \in G(d, CV, \fracτ{C})$ such that $L(M, P) \leq C ε.$ (b) There exists no $M \in G(d, V/C, Cτ)$ such that $L(M, P) \leq \fracε{C}.$ The answer is correct with probability at least $1-δ$ .

preprint2012arXiv

Damped random walks and the characteristic polynomial of the weighted Laplacian on a graph

For $λ>0$ , we define a $λ$ -damped random walk to be a random walk that is started from a random vertex of a graph and stopped at each step with probability $\fracλ{1+λ}$ , otherwise continued with probability $\frac{1}{1+λ}$ . We use the Aldous-Broder algorithm (\cite{aldous, broder}) of generating a random spanning tree and the Matrix-tree theorem to relate the values of the characteristic polynomial of the Laplacian at $\pm λ$ and the stationary measures of the sets of nodes visited by $i$ independent $λ$ -damped random walks for $i \in \N$ . As a corollary, we obtain a new characterization of the non-zero eigenvalues of the Weighted Graph Laplacian.

preprint2012arXiv

Random Trees in Electrical Networks

This paper contains results relating currents and voltages in resistive networks to appropriate random trees or forests in those networks.

preprint2010arXiv

Learning with Spectral Kernels and Heavy-Tailed Data

Two ubiquitous aspects of large-scale data analysis are that the data often have heavy-tailed properties and that diffusion-based or spectral-based methods are often used to identify and extract structure of interest. Perhaps surprisingly, popular distribution-independent methods such as those based on the VC dimension fail to provide nontrivial results for even simple learning problems such as binary classification in these two settings. In this paper, we develop distribution-dependent learning methods that can be used to provide dimension-independent sample complexity bounds for the binary classification problem in these two popular settings. In particular, we provide bounds on the sample complexity of maximum margin classifiers when the magnitude of the entries in the feature vector decays according to a power law and also when learning is performed with the so-called Diffusion Maps kernel. Both of these results rely on bounding the annealed entropy of gap-tolerant classifiers in a Hilbert space. We provide such a bound, and we demonstrate that our proof technique generalizes to the case when the margin is measured with respect to more general Banach space norms. The latter result is of potential interest in cases where modeling the relationship between data elements as a dot product in a Hilbert space is too restrictive.

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Institution

Affiliation not imported yet

This author record came from a source that does not expose affiliation metadata. Once the author claims the profile or we enrich the record from another provider, this section will link to the concrete institution.

Topic footprint

Fields this researcher appears in

Source provenance

Where this author record came from

arxivconfidence 95%

external id: arxiv:2605.02327:author:5:hariharan-narayanan

Imported May 20, 2026Synced May 21, 2026

Alexander Rakhlin3 works Charles Fefferman3 works Matti Lassas2 works Tengyuan Liang2 works

Hariharan Narayanan

Research overview

Is this your work?

15 publications

Denoising data using convex relaxations

Fitting a manifold of large reach to noisy data

On the mixing time of coordinate Hit-and-Run

Implicit Linear Algebra and Basic Circuit Theory

John's Walk

Random concave functions on an equilateral lattice with periodic Hessians I: entropy and Laplacians

Escaping the Local Minima via Simulated Annealing: Optimization of Approximately Convex Functions

Randomized Interior Point methods for Sampling and Optimization

On Zeroth-Order Stochastic Convex Optimization via Random Walks

Efficient Sampling from Time-Varying Log-Concave Distributions

Estimating Certain Non-Zero Littlewood-Richardson Coefficients

Testing the Manifold Hypothesis

Damped random walks and the characteristic polynomial of the weighted Laplacian on a graph

Random Trees in Electrical Networks

Learning with Spectral Kernels and Heavy-Tailed Data

See the researcher in context

Building this map preview