• A recent model for property testing of probability distributions (Chakraborty et al., ITCS 2013, Canonne et al., SICOMP 2015) enables tremendous savings in the sample complexity of testing algorithms, by allowing them to condition the sampling on subsets of the domain. In particular, Canonne, Ron, and Servedio (SICOMP 2015) showed that, in this setting, testing identity of an unknown distribution $D$ (whether $D=D^\ast$ for an explicitly known $D^\ast$) can be done with a constant number of queries, independent of the support size $n$ -- in contrast to the required $\Omega(\sqrt{n})$ in the standard sampling model. It was unclear whether the same stark contrast exists for the case of testing equivalence, where both distributions are unknown. While Canonne et al. established a $\mathrm{poly}(\log n)$-query upper bound for equivalence testing, very recently brought down to $\tilde O(\log\log n)$ by Falahatgar et al. (COLT 2015), whether a dependence on the domain size $n$ is necessary was still open, and explicitly posed by Fischer at the Bertinoro Workshop on Sublinear Algorithms (2014). We show that any testing algorithm for equivalence must make $\Omega(\sqrt{\log\log n})$ queries in the conditional sampling model. This demonstrates a gap between identity and equivalence testing, absent in the standard sampling model (where both problems have sampling complexity $n^{\Theta(1)}$). We also obtain results on the query complexity of uniformity testing and support-size estimation with conditional samples. We answer a question of Chakraborty et al. (ITCS 2013) showing that non-adaptive uniformity testing indeed requires $\Omega(\log n)$ queries in the conditional model. For the related problem of support-size estimation, we provide both adaptive and non-adaptive algorithms, with query complexities $\mathrm{poly}(\log\log n)$ and $\mathrm{poly}(\log n)$, respectively.
  • Independent samples from an unknown probability distribution $\mathbf{p}$ on a domain of size $k$ are distributed across $n$ players, with each player holding one sample. Each player can communicate $\ell$ bits to a central referee in a simultaneous message passing (SMP) model of communication to help the referee infer a property of the unknown $\mathbf{p}$. When $\ell\geq\log k$ bits, the problem reduces to the well-studied collocated case where all the samples are available in one place. In this work, we focus on the communication-starved setting of $\ell < \log k$, in which the landscape may change drastically. We propose a general formulation for inference problems in this distributed setting, and instantiate it to two prototypical inference questions: learning and uniformity testing.
  • We develop differentially private methods for estimating various distributional properties. Given a sample from a discrete distribution $p$, some functional $f$, and accuracy and privacy parameters $\alpha$ and $\varepsilon$, the goal is to estimate $f(p)$ up to accuracy $\alpha$, while maintaining $\varepsilon$-differential privacy of the sample. We prove almost-tight bounds on the sample size required for this problem for several functionals of interest, including support size, support coverage, and entropy. We show that the cost of privacy is negligible in a variety of settings, both theoretically and experimentally. Our methods are based on a sensitivity analysis of several state-of-the-art methods for estimating these properties with sublinear sample complexities.
  • We consider discrete distribution estimation over $k$ elements under $\varepsilon$-local differential privacy from $n$ samples. The samples are distributed across users who send privatized versions of their sample to the server. All previously known sample optimal algorithms require linear (in $k$) communication complexity in the high privacy regime $(\varepsilon<1)$, and have a running time that grows as $n\cdot k$, which can be prohibitive for large domain size $k$. We study the task simultaneously under four resource constraints, privacy, sample complexity, computational complexity, and communication complexity. We propose \emph{Hadamard Response (HR)}, a local non-interactive privatization mechanism with order optimal sample complexity (for all privacy regimes), a communication complexity of $\log k+2$ bits, and runs in nearly linear time. Our encoding and decoding mechanisms are based on Hadamard matrices, and are simple to implement. The gain in sample complexity comes from the large Hamming distance between rows of Hadamard matrices, and the gain in time complexity is achieved by using the Fast Walsh-Hadamard transform. We compare our approach with Randomized Response (RR), RAPPOR, and subset-selection mechanisms (SS), theoretically, and experimentally. For $k=10000$, our algorithm runs about 100x faster than SS, and RAPPOR.
  • We study the fundamental problems of identity testing (goodness of fit), and closeness testing (two sample test) of distributions over $k$ elements, under differential privacy. While the problems have a long history in statistics, finite sample bounds for these problems have only been established recently. In this work, we derive upper and lower bounds on the sample complexity of both the problems under $\varepsilon$-differential privacy. Our results improve over the best known algorithms for identity testing, and are the first results for differentially private closeness testing. Our bounds are tight up to a constant factor whenever the number of samples is $O(k)$, a regime which has garnered much attention over the last decade since it allows property testing even when only a fraction of the domain is observed. Our upper bounds are obtained by converting (and sometimes combining) the known non-private testing algorithms into differentially private algorithms. We propose a simple, yet general procedure based on coupling of distributions, to establish sample complexity lower bounds for differentially private algorithms.
  • Unlike compressive sensing where the measurement outputs are assumed to be real-valued and have infinite precision, in "one-bit compressive sensing", measurements are quantized to one bit, their signs. In this work, we show how to recover the support of sparse high-dimensional vectors in the one-bit compressive sensing framework with an asymptotically near-optimal number of measurements. We also improve the bounds on the number of measurements for approximately recovering vectors from one-bit compressive sensing measurements. Our results are universal, namely the same measurement scheme works simultaneously for all sparse vectors. Our proof of optimality for support recovery is obtained by showing an equivalence between the task of support recovery using 1-bit compressive sensing and a well-studied combinatorial object known as Union Free Families.
  • The advent of data science has spurred interest in estimating properties of distributions over large alphabets. Fundamental symmetric properties such as support size, support coverage, entropy, and proximity to uniformity, received most attention, with each property estimated using a different technique and often intricate analysis tools. We prove that for all these properties, a single, simple, plug-in estimator---profile maximum likelihood (PML)---performs as well as the best specialized techniques. This raises the possibility that PML may optimally estimate many other symmetric properties.
  • We study the fixed design segmented regression problem: Given noisy samples from a piecewise linear function $f$, we want to recover $f$ up to a desired accuracy in mean-squared error. Previous rigorous approaches for this problem rely on dynamic programming (DP) and, while sample efficient, have running time quadratic in the sample size. As our main contribution, we provide new sample near-linear time algorithms for the problem that -- while not being minimax optimal -- achieve a significantly better sample-time tradeoff on large datasets compared to the DP approach. Our experimental evaluation shows that, compared with the DP approach, our algorithms provide a convergence rate that is only off by a factor of $2$ to $4$, while achieving speedups of three orders of magnitude.
  • We study maximum selection and sorting of $n$ numbers using pairwise comparators that output the larger of their two inputs if the inputs are more than a given threshold apart, and output an adversarially-chosen input otherwise. We consider two adversarial models. A non-adaptive adversary that decides on the outcomes in advance based solely on the inputs, and an adaptive adversary that can decide on the outcome of each query depending on previous queries and outcomes. Against the non-adaptive adversary, we derive a maximum-selection algorithm that uses at most $2n$ comparisons in expectation, and a sorting algorithm that uses at most $2n \ln n$ comparisons in expectation. These numbers are within small constant factors from the best possible. Against the adaptive adversary, we propose a maximum-selection algorithm that uses $\Theta(n\log (1/{\epsilon}))$ comparisons to output a correct answer with probability at least $1-\epsilon$. The existence of this algorithm affirmatively resolves an open problem of Ajtai, Feldman, Hassadim, and Nelson. Our study was motivated by a density-estimation problem where, given samples from an unknown underlying distribution, we would like to find a distribution in a known class of $n$ candidate distributions that is close to underlying distribution in $\ell_1$ distance. Scheffe's algorithm outputs a distribution at an $\ell_1$ distance at most 9 times the minimum and runs in time $\Theta(n^2\log n)$. Using maximum selection, we propose an algorithm with the same approximation guarantee but run time of $\Theta(n\log n)$.
  • It was recently shown that estimating the Shannon entropy $H({\rm p})$ of a discrete $k$-symbol distribution ${\rm p}$ requires $\Theta(k/\log k)$ samples, a number that grows near-linearly in the support size. In many applications $H({\rm p})$ can be replaced by the more general R\'enyi entropy of order $\alpha$, $H_\alpha({\rm p})$. We determine the number of samples needed to estimate $H_\alpha({\rm p})$ for all $\alpha$, showing that $\alpha < 1$ requires a super-linear, roughly $k^{1/\alpha}$ samples, noninteger $\alpha>1$ requires a near-linear $k$ samples, but, perhaps surprisingly, integer $\alpha>1$ requires only $\Theta(k^{1-1/\alpha})$ samples. Furthermore, developing on a recently established connection between polynomial approximation and estimation of additive functions of the form $\sum_{x} f({\rm p}_x)$, we reduce the sample complexity for noninteger values of $\alpha$ by a factor of $\log k$ compared to the empirical estimator. The estimators achieving these bounds are simple and run in time linear in the number of samples. Our lower bounds provide explicit constructions of distributions with different R\'enyi entropies that are hard to distinguish.
  • Given samples from an unknown distribution $p$, is it possible to distinguish whether $p$ belongs to some class of distributions $\mathcal{C}$ versus $p$ being far from every distribution in $\mathcal{C}$? This fundamental question has received tremendous attention in statistics, focusing primarily on asymptotic analysis, and more recently in information theory and theoretical computer science, where the emphasis has been on small sample size and computational complexity. Nevertheless, even for basic properties of distributions such as monotonicity, log-concavity, unimodality, independence, and monotone-hazard rate, the optimal sample complexity is unknown. We provide a general approach via which we obtain sample-optimal and computationally efficient testers for all these distribution families. At the core of our approach is an algorithm which solves the following problem: Given samples from an unknown distribution $p$, and a known distribution $q$, are $p$ and $q$ close in $\chi^2$-distance, or far in total variation distance? The optimality of our testers is established by providing matching lower bounds with respect to both $n$ and $\varepsilon$. Finally, a necessary building block for our testers and an important byproduct of our work are the first known computationally efficient proper learners for discrete log-concave and monotone hazard rate distributions.
  • We design a new, fast algorithm for agnostically learning univariate probability distributions whose densities are well approximated by piecewise polynomial functions. Let $f$ be the density function of an arbitrary univariate distribution, and suppose that $f$ is $\mathrm{OPT}$-close in $L_1$-distance to an unknown piecewise polynomial function with $t$ interval pieces and degree $d$. Our algorithm draws $n = O(t(d+1)/\epsilon^2)$ samples from $f$, runs in time $\tilde{O}(n \cdot \mathrm{poly}(d))$, and with probability at least $9/10$ outputs an $O(t)$-piecewise degree-$d$ hypothesis $h$ that is $4 \cdot \mathrm{OPT} +\epsilon$ close to $f$. Our general algorithm yields (nearly) sample-optimal and nearly-linear time estimators for a wide range of structured distribution families over both continuous and discrete domains in a unified way. For most of our applications, these are the first sample-optimal and nearly-linear time estimators in the literature. As a consequence, our work resolves the sample and computational complexities of a broad class of inference tasks via a single "meta-algorithm". Moreover, we experimentally demonstrate that our algorithm performs very well in practice. Our algorithm consists of three "levels": (i) At the top level, we employ an iterative greedy algorithm for finding a good partition of the real line into the pieces of a piecewise polynomial. (ii) For each piece, we show that the sub-problem of finding a good polynomial fit on the current interval can be solved efficiently with a separation oracle method. (iii) We reduce the task of finding a separating hyperplane to a combinatorial problem and give an efficient algorithm for this problem. Combining these three procedures gives a density estimation algorithm with the claimed guarantees.
  • A Poisson Binomial distribution over $n$ variables is the distribution of the sum of $n$ independent Bernoullis. We provide a sample near-optimal algorithm for testing whether a distribution $P$ supported on $\{0,...,n\}$ to which we have sample access is a Poisson Binomial distribution, or far from all Poisson Binomial distributions. The sample complexity of our algorithm is $O(n^{1/4})$ to which we provide a matching lower bound. We note that our sample complexity improves quadratically upon that of the naive "learn followed by tolerant-test" approach, while instance optimal identity testing [VV14] is not applicable since we are looking to simultaneously test against a whole family of distributions.
  • The Poisson-sampling technique eliminates dependencies among symbol appearances in a random sequence. It has been used to simplify the analysis and strengthen the performance guarantees of randomized algorithms. Applying this method to universal compression, we relate the redundancies of fixed-length and Poisson-sampled sequences, use the relation to derive a simple single-letter formula that approximates the redundancy of any envelope class to within an additive logarithmic term. As a first application, we consider i.i.d. distributions over a small alphabet as a step-envelope class, and provide a short proof that determines the redundancy of discrete distributions over a small al- phabet up to the first order terms. We then show the strength of our method by applying the formula to tighten the existing bounds on the redundancy of exponential and power-law classes, in particular answering a question posed by Boucheron, Garivier and Gassiat.
  • Motivated by mass-spectrometry protein sequencing, we consider a simply-stated problem of reconstructing a string from the multiset of its substring compositions. We show that all strings of length 7, one less than a prime, or one less than twice a prime, can be reconstructed uniquely up to reversal. For all other lengths we show that reconstruction is not always possible and provide sometimes-tight bounds on the largest number of strings with given substring compositions. The lower bounds are derived by combinatorial arguments and the upper bounds by algebraic considerations that precisely characterize the set of strings with the same substring compositions in terms of the factorization of bivariate polynomials. The problem can be viewed as a combinatorial simplification of the turnpike problem, and its solution may shed light on this long-standing problem as well. Using well known results on transience of multi-dimensional random walks, we also provide a reconstruction algorithm that reconstructs random strings over alphabets of size $\ge4$ in optimal near-quadratic time.
  • Statistical and machine-learning algorithms are frequently applied to high-dimensional data. In many of these applications data is scarce, and often much more costly than computation time. We provide the first sample-efficient polynomial-time estimator for high-dimensional spherical Gaussian mixtures. For mixtures of any $k$ $d$-dimensional spherical Gaussians, we derive an intuitive spectral-estimator that uses $\mathcal{O}_k\bigl(\frac{d\log^2d}{\epsilon^4}\bigr)$ samples and runs in time $\mathcal{O}_{k,\epsilon}(d^3\log^5 d)$, both significantly lower than previously known. The constant factor $\mathcal{O}_k$ is polynomial for sample complexity and is exponential for the time complexity, again much smaller than what was previously known. We also show that $\Omega_k\bigl(\frac{d}{\epsilon^2}\bigr)$ samples are needed for any algorithm. Hence the sample complexity is near-optimal in the number of dimensions. We also derive a simple estimator for one-dimensional mixtures that uses $\mathcal{O}\bigl(\frac{k \log \frac{k}{\epsilon} }{\epsilon^2} \bigr)$ samples and runs in time $\widetilde{\mathcal{O}}\left(\bigl(\frac{k}{\epsilon}\bigr)^{3k+1}\right)$. Our other technical contributions include a faster algorithm for choosing a density estimate from a set of distributions, that minimizes the $\ell_1$ distance to an unknown underlying distribution.
  • This paper proposes a novel method for segmentation of images by hierarchical multilevel thresholding. The method is global, agglomerative in nature and disregards pixel locations. It involves the optimization of the ratio of the unbiased estimators of within class to between class variances. We obtain a recursive relation at each step for the variances which expedites the process. The efficacy of the method is shown in a comparison with some well-known methods.
  • A novel algorithm is proposed for segmenting an image into multiple levels using its mean and variance. Starting from the extreme pixel values at both ends of the histogram plot, the algorithm is applied recursively on sub-ranges computed from the previous step, so as to find a threshold level and a new sub-range for the next step, until no significant improvement in image quality can be achieved. The method makes use of the fact that a number of distributions tend towards Dirac delta function, peaking at the mean, in the limiting condition of vanishing variance. The procedure naturally provides for variable size segmentation with bigger blocks near the extreme pixel values and finer divisions around the mean or other chosen value for better visualization. Experiments on a variety of images show that the new algorithm effectively segments the image in computationally very less time.