We study the problem of active nonparametric sequential two-sample testing over multiple heterogeneous data sources. In each time slot, a decision-maker adaptively selects one of $K$ data sources and receives a paired sample generated from that source for testing. The goal is to decide as quickly as possible whether the pairs are generated from the same distribution or not. The gain achieved by such adaptive sampling (in terms of smaller expected stopping time or larger error exponents) has been well-characterized for parametric models via Chernoff's adaptive MLE selection rule [1]. However, analogous results are not known for the case of nonparametric problems, such as two-sample testing, where we place no restrictions on the distributions. Our main contribution is a general active nonparametric testing procedure that combines an adaptive source-selecting strategy within the testing-by-betting framework of [2] that works under minimal distributional assumptions. In each time slot, our scheme proceeds by selecting a source according to a probability that mixes exploitation, favoring sources with the largest empirical distinguishability, and exploration via a vanishing greedy strategy. The (paired) observations so collected are then used to update the "betting-wealth process", which is a stochastic process guaranteed to be a nonnegative martingale under the null. The procedure stops and rejects the null when the wealth process exceeds an appropriate threshold; an event that is unlikely under the null. We show that our test controls the type-I error at a prespecified level-$α$ under the null, and establish its power-one property and a bound on its expected sample size under the alternative. Our results provide a precise characterization of the improvements achievable by a principled adaptive sampling strategy over its passive analog.

Lemma 2of Shekhar and Ramdas (2024), which was used to derive the upper bound on the expected stopping time stated in(12), contains an error. In this note, we fix this error and provide the correct justification of(12), whose expression remains unchanged up to small constants.

We propose a general framework for constructing powerful, sequential hypothesis tests for a large class of nonparametric testing problems. The null hypothesis for these problems is defined in an abstract form using the action of two known operators on the data distribution. This abstraction allows for a unified treatment of several classical tasks, such as two-sample testing, independence testing, and conditional-independence testing, as well as modern problems, such as testing for adversarial robustness of machine learning (ML) models. Our proposed framework has the following advantages over classical batch tests: 1) it continuously monitors online data streams and efficiently aggregates evidence against the null, 2) it provides tight control over the type I error without the need for multiple testing correction, 3) it adapts the sample size requirement to the unknown hardness of the problem. We develop a principled approach of leveraging the representation capability of ML models within the testing-by-betting framework, a game-theoretic approach for designing sequential tests. Empirical results on synthetic and real-world datasets demonstrate that tests instantiated using our general framework are competitive against specialized baselines on several tasks.

We consider the problem of sequential change detection under minimal assumptions on the distribution generating the stream of observations. Formally, our goal is to design a scheme for detecting any changes in a parameter or functional $\theta$ of the data stream distribution that has small detection delay, but guarantees control on the frequency of false alarms in the absence of changes. We describe a simple reduction from sequential change detection to sequential estimation using confidence sequences (CSs): begin a new level-$(1-\alpha)$ CS at each time step, and proclaim a change as soon as the intersection of all active CSs becomes empty. We prove that the average run length of our scheme is at least $1/\alpha$, resulting in a change detection scheme with minimal structural assumptions (thus allowing for possibly dependent observations, and nonparametric distribution classes), but strong guarantees. We also describe an interesting parallel with Lorden's reduction from change detection to sequential testing and connections to the recent ''e-detector'' framework.

We study the problem of designing consistent sequential two-sample tests in a nonparametric setting. Guided by the principle of testing by betting, we reframe this task into that of selecting a sequence of payoff functions that maximize the wealth of a fictitious bettor, betting against the null in a repeated game. In this setting, the relative increase in the bettor’s wealth has a precise interpretation as the measure of evidence against the null, and thus our sequential test rejects the null when the wealth crosses an appropriate threshold. We develop a general framework for setting up the betting game for two-sample testing, in which the payoffs are selected by a prediction strategy as data-driven predictable estimates of the witness function associated with the variational representation of some statistical distance measures, such as integral probability metrics (IPMs). We then formally relate the statistical properties of the test (such as consistency, type-II error exponent and expected sample size) to the regret of the corresponding prediction strategy. We construct a practical sequential two-sample test by instantiating our general strategy with the kernel-MMD metric, and demonstrate its ability to adapt to the difficulty of the unknown alternative through theoretical and empirical results. Our framework is versatile, and easily extends to other problems; we illustrate this by applying our approach to construct consistent tests for the following problems: (i) time-varying two-sample testing with non-exchangeable observations, and (ii) an abstract class of “invariant” testing problems, including symmetry and independence testing.

We present a simple reduction from sequential estimation to sequential changepoint detection (SCD). In short, suppose we are interested in detecting changepoints in some parameter or functional $\theta$ of the underlying distribution. We demonstrate that if we can construct a confidence sequence (CS) for $\theta$, then we can also successfully perform SCD for $\theta$. This is accomplished by checking if two CSs — one forwards and the other backwards — ever fail to intersect. Since the literature on CSs has been rapidly evolving recently, the reduction provided in this paper immediately solves several old and new change detection problems. Further, our “backward CS”, constructed by reversing time, is new and potentially of independent interest. We provide strong nonasymptotic guarantees on the frequency of false alarms and detection delay, and demonstrate numerical effectiveness on several problems.

We introduce the notion of risk-limiting financial audits (RLFA): procedures that manually evaluate a subset of $N$ financial transactions to check the validity of a claimed assertion $\mathcal{A}$ about the transactions. More specifically, RLFA satisfy two properties: (i) if $\mathcal{A}$ is false, they correctly disprove it with probability at least $1-\delta$, and (ii) they validate the correctness of $\mathcal{A}$ with probability $1$, if it is true. We propose a general RLFA strategy, by constructing new confidence sequences (CSs) for the weighted average of $N$ unknown values, based on samples drawn without replacement from a (randomized) weighted sampling scheme. Next, we develop methods to improve the quality of CSs by incorporating side information about the unknown values. We show that when the side information is sufficiently accurate, it can directly drive the sampling. For the case where the accuracy is unknown a priori, we introduce an alternative approach using control variates. Crucially, our construction adapts to the quality of side information by strongly leveraging the side information if it is highly predictive, and learning to ignore it if it is uninformative. Our methods also recover the state-of-the-art bounds for the special case of uniformly sampled observations with no side information, which has already found applications in election auditing. The harder weighted case with general side information solves the more challenging problem of AI-assisted financial auditing.

In nonparametric independence testing, we observe i.i.d.\ data $\{(X_i,Y_i)\}_{i=1}^n$, where $X \in \mathcal{X}, Y \in \mathcal{Y}$ lie in any general spaces, and we wish to test the null that $X$ is independent of $Y$. Modern test statistics such as the kernel Hilbert--Schmidt Independence Criterion (HSIC) and Distance Covariance (dCov) have intractable null distributions due to the degeneracy of the underlying U-statistics. Hence, in practice, one often resorts to using permutation testing, which provides a nonasymptotic guarantee at the expense of recalculating the quadratic-time statistics (say) a few hundred times. In this paper, we provide a simple but nontrivial modification of HSIC and dCov (called xHSIC and xdCov, pronounced “cross” HSIC/dCov) so that they have a limiting Gaussian distribution under the null, and thus do not require permutations. We show that our new tests, like the originals, are consistent against fixed alternatives, and minimax rate optimal against smooth local alternatives. Numerical simulations demonstrate that compared to the permutation tests, our variants have the same power within a constant factor, giving practitioners a new option for large problems or data-analysis pipelines where computation, not sample size, could be the bottleneck.

We study the problem of designing an adaptive strategy for querying a noisy zeroth-order-oracle to efficiently learn about the optimizer of an unknown function $f$. To make the problem tractable, we assume that $f$ lies in the reproducing kernel Hilbert space (RKHS) associated with a known kernel $K$, with its norm bounded by $M<\infty$. Prior results, working in a minimax framework, have characterized the worst-case (over all functions in the problem class) limits on regret achievable by any algorithm, and have constructed algorithms with matching (modulo polylogarithmic factors) worst-case performance for the Matern family of kernels. These results suffer from two drawbacks. First, the minimax lower bound gives limited information about the limits of regret achievable by commonly used algorithms on a specific problem instance $f$. Second, the existing upper bound analysis fails to adapt to easier problem instances within the function class. Our work takes steps to address both these issues. First, we derive instance-dependent regret lower bounds for algorithms with uniformly (over the function class) vanishing normalized cumulative regret. Our result, valid for several practically relevant kernelized bandits algorithms, such as, GP-UCB, GP-TS and SupKernelUCB, identifies a fundamental complexity measure associated with every problem instance. We then address the second issue, by proposing a new minimax near-optimal algorithm that also adapts to easier problem instances.

Consider the problem of optimizing a black-box function under the assumption that the function is Holder smooth and has bounded norm in the reproducing kernel Hilbert space associated with a given kernel. We propose the LP-GP-UCB algorithm which augments a Gaussian process surrogate model with local polynomial estimators of the function to construct a multi-scale upper confidence bound to guide the search for the optimizer. We provide high probability bounds on the cumulative regret in terms of the maximum information gain and smoothness parameters for the kernel. We then show that the Holder smoothness assumption is satisfied for several commonly used and practically relevant kernels—the Matern, rational- quadratic, γ-exponential, and piecewise-polynomial kernels—and obtain explicit regret bounds for them as a result. These regret bounds establish the near-optimality of LP-GP-UCB for these kernels and are also the first explicit bounds for many of them. Finally, we demonstrate the practical benefits experimentally.

The kernel Maximum Mean Discrepancy~(MMD) is a popular multivariate distance metric between distributions. The usual kernel-MMD test statistic (for two-sample testing) is a degenerate U-statistic under the null, and thus it has an intractable limiting null distribution. Hence, the standard approach for designing a level-$(1-\alpha)$ two-sample test using this statistic involves selecting the rejection threshold as the $(1-\alpha)$-quantile of the permutation distribution. The resulting nonparametric test has finite-sample validity but suffers from large computational cost, since the test statistic must be recomputed for every permutation. We propose the cross-MMD, a new quadratic time MMD test statistic based on sample-splitting and studentization. We prove that under mild assumptions, it has a standard normal limiting distribution under the null. Importantly, we also show that the resulting test is consistent against any fixed alternative, and when using the Gaussian kernel, it has minimax rate-optimal power against local alternatives. For large sample-sizes, our new cross-MMD provides a significant speedup over the MMD, for only a slight loss in power.

We consider the problem of Bayesian Optimization (BO) in which the goal is to design an adaptive querying strategy to optimize a function $f:[0,1]^d\mapsto \reals$. The function is assumed to be drawn from a Gaussian Process, and can only be accessed through noisy oracle queries. The most commonly used oracle in BO literature is the noisy Zeroth-Order-Oracle (ZOO) which returns noise-corrupted function value $y = f(x) + \eta$ at any point $x \in \domain$ queried by the agent. A less studied oracle in BO is the First-Order-Oracle (FOO) which also returns noisy gradient value at the queried point. In this paper we consider the fundamental question of quantifying the possible improvement in regret that can be achieved under FOO access as compared to the case in which only ZOO access is available. Under some regularity assumptions on $K$, we first show that the expected cumulative regret with ZOO of any algorithm must satisfy a lower bound of $\Omega(\sqrt{2^d n})$, where $n$ is the query budget. This lower bound captures the appropriate scaling of the regret on both dimension $d$ and budget $n$, and relies on a novel reduction from BO to a multi-armed bandit (MAB) problem. We then propose a two-phase algorithm which, with some additional prior knowledge, achieves a vastly improved $\mc{O}\lp d (\log n)^2 \rp$ regret when given access to a FOO. Together, these two results highlight the significant value of incorporating gradient information in BO algorithms.

Machine learning models trained on uncurated datasets can often end up adversely affecting inputs belonging to underrepresented groups. To address this issue, we consider the problem of adaptively constructing training sets which allow us to learn classifiers that are fair in a {\em minimax} sense. We first propose an adaptive sampling algorithm based on the principle of \emph{optimism}, and derive theoretical bounds on its performance. We also propose heuristic extensions of this algorithm suitable for application to large scale, practical problems. Next, by deriving algorithm independent lower-bounds for a specific class of problems, we show that the performance achieved by our adaptive scheme cannot be improved in general. We then validate the benefits of adaptively constructing training sets via experiments on synthetic tasks with logistic regression classifiers, as well as on several real-world tasks using convolutional neural networks (CNNs).

We consider the problem of allocating a fixed budget of samples to a finite set of discrete distributions to learn them uniformly well (minimizing the maximum error) in terms of four common distance measures: $\ell_2^2$, $\ell_1$, $f$-divergence, and separation distance. To present a unified treatment of these distances, we first propose a general \emph{optimistic tracking algorithm} and analyze its sample allocation performance w.r.t. an oracle. We then instantiate this algorithm for the four distance measures and derive bounds on their regret. We also show that the allocation performance of the proposed algorithm cannot, in general, be improved, by deriving lower-bounds on the expected deviation from the oracle allocation for any adaptive scheme. We verify our theoretical findings through some experiments. Finally, we show that the techniques developed in the paper can be easily extended to learn some classes of continuous distributions as well as to the related setting of minimizing the average error (in terms of the four distances) in learning a set of distributions.

We consider the problem of binary classification with the caveat that the learner can abstain from declaring a label incurring a cost λ ∈ [0,1/2] in the process. This is referred to as the problem of binary classification with a fixed-cost of abstention. For this problem, we propose an active learning strategy that constructs a non-uniform partition of the input space and focuses sampling in the regions near the decision boundaries. Our proposed algorithm can work in all the commonly used active learning query models, namely membership-query, pool-based and stream-based. We obtain an upper bound on the excess risk of our proposed algorithm under standard smoothness and margin assumptions and demonstrate its minimax near-optimality by deriving a matching (modulo poly-logarithmic factors) lower bound. The achieved minimax rates are always faster than the corresponding rates in the passive setting, and furthermore the improvement increases with larger values of the smoothness and margin parameters.

We study the problem of efficient exploration in order to learn an accurate model of an environment, modeled as a Markov decision process (MDP). Efficient exploration in this problem requires the agent to identify the regions in which estimating the model is more difficult and then exploit this knowledge to collect more samples there. In this paper, we formalize this problem, introduce the first algorithm to learn an $\epsilon$-accurate estimate of the dynamics, and provide its sample complexity analysis. While this algorithm enjoys strong guarantees in the large-sample regime, it tends to have a poor performance in early stages of exploration. To address this issue, we propose an algorithm that is based on maximum weighted entropy, a heuristic that stems from common sense and our theoretical analysis. The main idea here is to cover the entire state-action space with the weight proportional to the noise in their transition functions. Using a number of simple domains with heterogeneous noise in their transitions, we show that our heuristic-based algorithm outperforms both our original algorithm and the maximum entropy algorithm in the small sample regime, while achieving similar asymptotic performance as that of the original algorithm.

In this paper, the problem of estimating the level set of a black-box function from noisy and expensive evaluation queries is considered. A new algorithm for this problem in the Bayesian framework with a Gaussian Process (GP) prior is proposed. The proposed algorithm employs a hierarchical sequence of partitions to explore different regions of the search space at varying levels of detail depending upon their proximity to the level set boundary. It is shown that this approach results in the algorithm having a low complexity implementation whose computational cost is significantly smaller than the existing algorithms for higher dimensional search space $\X$. Furthermore, high probability bounds on a measure of discrepancy between the estimated level set and the true level set for the the proposed algorithm are obtained, which are shown to be strictly better than the existing guarantees for a large class of GPs.In the process, a tighter characterization of the information gain of the proposed algorithm is obtained which takes into account the structured nature of the evaluation points. This approach improves upon the existing technique of bounding the information gain with maximum information gain.

Species tree reconstruction from genomic data is increasingly performed using methods that account for sources of gene tree discordance such as incomplete lineage sorting. One popular method for reconstructing species trees from unrooted gene tree topologies is ASTRAL. In this paper, we derive theoretical sample complexity results for the number of genes required by ASTRAL to guarantee reconstruction of the correct species tree with high probability. We also validate those theoretical bounds in a simulation study. Our results indicate that ASTRAL requires gene trees to reconstruct the species tree correctly with high probability where is the number of species and is the length of the shortest branch in the species tree. Our simulations, some under the anomaly zone, show trends consistent with the theoretical bounds and also provide some practical insights on the conditions where ASTRAL works well.