arXiv subjects

Sourabh Bhattacharya

Publications and source records attributed to Sourabh Bhattacharya.

At least 19 recordsLinked to original sources

arXiv2026

Function Optimization with Posterior Gaussian Derivative Process

We propose a Bayesian optimization algorithm for differentiable functions with available first and second partial derivatives. The objective function is modeled as a Gaussian process, inducing a Gaussian derivative process. Given initial function evaluations, we obtain the posterior of the derivative process and construct a posterior for stationary points by setting the derivative to zero. A recursive importance-resampling scheme with prior constraints on the gradient norm and Hessian definiteness drives the algorithm toward the true optima. We prove almost sure convergence as the number of stages tends to infinity, relying on novel results establishing almost sure uniform convergence of the Gaussian process and its derivative process posteriors to the true function and its derivatives under fixed-domain infill asymptotics; rates of convergence are also derived. We also give a Bayesian characterization of the number of optima. The method is demonstrated on five diverse problems, including finding maxima, minima, saddle points, and inconclusive cases, ranging from one-dimensional to 100-dimensional nonlinear least-squares. On a real-world Poisson regression (AIDS deaths data), our procedure achieves a substantially smaller gradient norm at the MLE than Fisher scoring, BFGS, simulated annealing, and multi-start quasi-Newton. The posterior simulation nature allows exploration of neighborhoods of solutions from other methods, yielding more accurate results. Code is available at https://github.com/Sourabh-Bhattacharya/FUNCTION_OPT_GDP.

math.OC↗

arXiv2026

The Bayesian Reflex: A Predictive Coding Engine for Artificial Intelligence

Predictive coding offers a powerful theory of cortical computation, but corresponding scalable algorithmic implementations for artificial intelligence have remained elusive. This paper introduces the Bayesian reflex, a computational framework that directly instantiates predictive coding through three pillars: belief maintenance via hierarchical generative models, sequential Bayesian updating via prediction-error minimization, and uncertainty-driven action via active inference. We show that recent breakthroughs---ellipsoidal decomposition for exact $i.i.d.$ sampling, recursive Gaussian processes for deep hierarchical inference, and derivative-aware Bayesian optimization---provide the missing algorithmic ingredients. The resulting framework enables mathematically principled, scalable, and brain-inspired continual learning, perception, and decision-making. We illustrate its versatility through applications ranging from climate model evaluation to prime number discovery, offering a blueprint for truly adaptive artificial intelligence.

cs.AI↗

arXiv2026

Recursive Gaussian Processes and the Bayesian Brain

Predictive coding offers a powerful framework for cortical computation, yet scalable implementations that respect both Bayesian exactness and neurobiological constraints remain scarce. We bridge this gap by formally connecting predictive coding to Recursive Gaussian Processes (RGPs). RGPs employ a single Gaussian process $ g(t, \cdot) $ indexed by layer index and input value, preventing the representational collapse of standard deep Gaussian processes while allowing learnable cross-layer dependence via $ r_{1g} $. We demonstrate that RGPs intrinsically implement hierarchical Bayesian inference, uncertainty propagation, and precision-weighted prediction error. Critically, we map RGP components---the shared GP, spike-and-slab variable selection, and MCMC dynamics---onto the canonical cortical microcircuit, providing a neurobiological substrate for these computations. Drawing on the free energy principle, we show that RGP inference minimizes variational free energy, formally linking Bayesian mechanics to neuronal dynamics. Our synthesis positions RGPs as both a principled computational tool and a candidate model for the brain's predictive machinery, generating testable predictions for laminar-specific dynamics and spectral asymmetries between feedforward and feedback processing.

q-bio.NC↗

arXiv2026

The Bayesian Reflex: Online Learning as the Autonomic Nervous System of Modern and Future AI

This chapter introduces the Bayesian reflex -- an analogy with the autonomic nervous system -- as a unifying framework for online learning in AI. Bayesian online algorithms automatically maintain equilibrium in dynamic environments via three mechanisms: belief maintenance through probabilistic representations, sequential updating via Bayes' theorem, and uncertainty-driven action balancing exploration and exploitation. We survey online Bayesian methods, highlighting two computational principles: the look-up table principle for sequential inference in function space, and the ellipsoidal decomposition framework for nearly exact i.i.d. sampling from arbitrary posteriors. These principles are generalized across dynamic emulation, nonparametric state-space models, circular time series, inverse regression for climate model evaluation, and deep architectures via Recursive Gaussian Processes. Decision-making is explored via Thompson sampling and restless bandits. We extend the framework to assess infinite series convergence (applied to climate dynamics and the Riemann Hypothesis), model prime number distributions leading to the discovery of 184 strong Mersenne prime candidates, detect stationarity, and characterize point processes. The Bayesian reflex provides a foundational infrastructure for adaptive AI that continuously learns in a complex world.

stat.ME↗

arXiv2026

How Ominous is the Premonition of Future Global Warming?

Global warming, the phenomenon of increasing global average temperature in the recent decades, is receiving wide attention due to its very significant adverse effects on climate. Whether global warming will continue even in the future, is a question that is most important to investigate. In this regard, the so-called general circulation models (GCMs) have attempted to project the future climate, and nearly all of them exhibit alarming rates of global temperature rise in the future. Although global warming in the current time frame is undeniable, it is important to assess the validity of the future predictions of the GCMs. In this article, we attempt such a study using our recently-developed Bayesian multiple testing paradigm for model selection in inverse regression problems. The model we assume for the global temperature time series is based on Gaussian process emulation of the black box scenario, realistically treating the dynamic evolution of the time series as unknown. We apply our ideas to datasets available from the Intergovernmental Panel on Climate Change (IPCC) website. The best GCM models selected by our method under different assumptions on future climate change scenarios do not convincingly support the present global warming pattern when only the future predictions are considered known. Using our Gaussian process idea, we also forecast the future temperature time series given the current one. Interestingly, our results do not support drastic future global warming predicted by almost all the GCM models.

stat.AP↗

arXiv2026

IID Sampling from Intractable Distributions

We propose a novel methodology for drawing iid realizations from any target distribution on the Euclidean space with arbitrary dimension. No assumption of compact support is necessary for the validity of our theory and method. Our idea is to construct an appropriate infinite sequence of concentric closed ellipsoids, represent the target distribution as an infinite mixture on the central ellipsoid and the ellipsoidal annuli, and to construct efficient perfect samplers for the mixture components. In contrast with most of the existing works on perfect sampling, ours is not only a theoretically valid method, it is practically applicable to all target distributions on any dimensional Euclidean space and very much amenable to parallel computation. We validate the practicality and usefulness of our methodology by generating 10000 iid realizations from the standard distributions such as normal, Student's t with 5 degrees of freedom and Cauchy, for dimensions d = 1, 5, 10, 50, 100, as well as from a 50-dimensional mixture normal distribution. The implementation time in all the cases are very reasonable, and often less than a minute in our parallel implementation. The results turned out to be highly accurate. We also apply our method to draw 10000 iid realizations from the posterior distributions associated with the well-known Challenger data, a Salmonella data and the 160-dimensional challenging spatial example of the radionuclide count data on Rongelap Island. Again, we are able to obtain quite encouraging results with very reasonable computing time.

stat.CO↗

arXiv2026

Bayesian Nonparametrics for Gene-Gene and Gene-Environment Interactions in Case-Control Studies: A Synthesis and Extension

Gene-gene and gene-environment interactions are widely believed to play significant roles in explaining the variability of complex traits. While substantial research exists in this area, a comprehensive statistical framework that addresses multiple sources of uncertainty simultaneously remains lacking. In this article, we synthesize and propose extension of a novel class of Bayesian nonparametric approaches that account for interactions among genes, loci, and environmental factors while accommodating uncertainty about population substructure. Our contribution is threefold: (1) We provide a unified exposition of hierarchical Bayesian models driven by Dirichlet processes for genetic interactions, clarifying their conceptual advantages over traditional regression approaches; (2) We shed light on new computational strategies that combine transformation-based MCMC with parallel processing for scalable inference; and (3) We present enhanced hypothesis testing procedures for identifying disease-predisposing loci.Through applications to myocardial infarction data, we demonstrate how these methods offer biological insights not readily obtainable from standard approaches. Our synthesis highlights the advantages of Bayesian nonparametric thinking in genetic epidemiology while providing practical guidance for implementation.

stat.ME↗

arXiv2026

A New Spatio-Temporal Model Exploiting Hamiltonian Equations

The solutions of Hamiltonian equations are known to describe the underlying phase space of a mechanical system. In this article, we propose a novel spatio-temporal model using a strategic modification of the Hamiltonian equations, incorporating appropriate stochasticity via Gaussian processes. The resultant spatio-temporal process, continuously varying with time, turns out to be nonparametric, non-stationary, non-separable, and non-Gaussian. Additionally, the lagged correlations converge to zero as the spatio-temporal lag goes to infinity. We investigate the theoretical properties of the new spatio-temporal process, including its continuity and smoothness properties. We derive methods for complete Bayesian inference using MCMC techniques in the Bayesian paradigm. The performance of our method has been compared with that of a non-stationary Gaussian process (GP) using two simulation studies, where our method shows a significant improvement over the non-stationary GP. Further, applying our new model to two real data sets revealed encouraging performance.

stat.ME↗

arXiv2026

Mean-Square Convergence of a New Parameterized Leapfrog Scheme for Hamiltonian Systems Driven by Gaussian Process Potentials

This paper establishes the mean-square convergence of a new stochastic, parameterized leapfrog scheme introduced in our companion paper Mazumder et al. (2026) for Hamiltonian systems with Gaussian process potentials. We consider a one-step numerical integrator and provide a complete, rigorous analysis under minimal regularity assumptions on the Gaussian potential. The key technical contribution is identifying and exploiting the symplectic structure ingrained in our stochastic, parameterized leapfrog method. Combined with local truncation error analysis, this leads to a global error bound of O(δt) in mean-square sense. Our results establish that although the spatio-temporal model of Mazumder et al. (2026) arises as the anticipated new stochastic leapfrog solution of a system of modified (parameterized) stochastic Hamiltonian equations, the new stochastic leapfrog actually solves the traditional stochastic Hamiltonian equations, driven by Gaussian process potential.

math.ST↗

arXiv2026

MPL-HMC: A Tunable Parameterized Leapfrog Framework for Robust Hamiltonian Monte Carlo

This article introduces the Modified Parameterized Leapfrog Hamiltonian Monte Carlo (MPL-HMC) method, a novel extension of HMC addressing key limitations through tunable integration parameters $α(δt)$ and $β(δt)$, enabling controlled perturbations to Hamiltonian dynamics. Theoretical analysis demonstrates MPL-HMC maintains approximate detailed balance. Extensive empirical evaluation reveals systematic performance improvements. The damping variant ($α_2=-0.1$, $β_2=-0.05$) achieves a 14-fold increase in effective sample size for Neal's funnel and 27\% better efficiency for pharmacokinetic models. The anti-damping variant ($α_2=0.1$, $β_2=0.05$) achieves $\hat{R}=1.026$ for Bayesian neural networks versus $\hat{R}=1.981$ for standard HMC. We introduce aggressive MPL-HMC for multimodal distributions, employing extreme parameters ($α_2=8.0$--$15.0$, $β_2=5.0$--$8.0$) with enhanced sampling to achieve full mode exploration where standard methods fail. All variants maintain computational efficiency identical to standard HMC while providing systematic control over damping, exploration, stability, and accuracy. The article provides rigorous mathematical foundations, implementation specifications, parameter tuning strategies, and comprehensive performance comparisons, extending HMC's applicability to previously challenging domains.

stat.CO↗

arXiv2025

Bayes Meets Riemann Again: Large Prime Discovery and Re-emergence of the Bone of Contention

Prime numbers have fascinated mathematicians since antiquity, with ongoing efforts to uncover both their properties and ever-larger examples. While giant primes rarely aid cryptography, they find use in areas such as locally decodable codes. Large prime-hunting, often brute-force in nature, is conceptually linked to the Riemann Hypothesis and the prime number theorem, which portrays prime distribution as essentially random. This motivates a statistical perspective, with Bayesian methodology providing a natural foundation. We show that the prime number theorem suggests a nonhomogeneous Poisson process for prime counts, yielding primes as waiting times. This process agrees with the prime number theorem, asymptotic results, and prime gap properties. Building on it, we develop a recursive Bayesian theory for large prime prediction and Riemann Hypothesis validation. The approach matches traditional but computationally infeasible non-recursive Bayesian formulations in the limit, and it strongly falsifies the Riemann Hypothesis. Finally, we propose a computational method using Transformation-based MCMC to simulate recursive posterior predictives. A simple change of variable enables simulation of Mersenne prime exponents. With modest computing resources, we identified 259 primes over 140 million, including 184 strong Mersenne candidates corresponding to potential primes with 42--242 million digits.

math.GM↗

arXiv2024

Bayesian Levy-Dynamic Spatio-Temporal Process: Towards Big Data Analysis

In this era of big data, all scientific disciplines are evolving fast to cope up with the enormity of the available information. So is statistics, the queen of science. Big data are particularly relevant to spatio-temporal statistics, thanks to much-improved technology in satellite based remote sensing and Geographical Information Systems. However, none of the existing approaches seem to meet the simultaneous demand of reality emulation and cheap computation. In this article, with the Levy random fields as the starting point, e construct a new Bayesian nonparametric, nonstationary and nonseparable dynamic spatio-temporal model with the additional realistic property that the lagged spatio-temporal correlations converge to zero as the lag tends to infinity. Although our Bayesian model seems to be intricately structured and is variable-dimensional with respect to each time index, we are able to devise a fast and efficient parallel Markov Chain Monte Carlo (MCMC) algorithm for Bayesian inference. Our simulation experiment brings out quite encouraging performance from our Bayesian Levy-dynamic approach. We finally apply our Bayesian Levy-dynamic model and methods to a sea surface temperature dataset consisting of 139,300 data points in space and time. Although not big data in the true sense, this is a large and highly structured data by any standard. Even for this large and complex data, our parallel MCMC algorithm, implemented on 80 processors, generated 110,000 MCMC realizations from the Levy-dynamic posterior within a single day, and the resultant Bayesian posterior predictive analysis turned out to be encouraging. Thus, it is not unreasonable to expect that with significantly more computing resources, it is feasible to analyse terabytes of spatio-temporal data with our new model and methods.

stat.ME↗

arXiv2024

Increasing Domain Infill Asymptotics for Stochastic Differential Equations Driven by Fractional Brownian Motion

Although statistical inference in stochastic differential equations (SDEs) driven by Wiener process has received significant attention in the literature, inference in those driven by fractional Brownian motion seem to have seen much less development in comparison, despite their importance in modeling long range dependence. In this article, we consider both classical and Bayesian inference in such fractional Brownian motion based SDEs. In particular, we consider asymptotic inference for two parameters in this regard; a multiplicative parameter associated with the drift function, and the so-called "Hurst parameter" of the fractional Brownian motion, when the time domain tends to infinity. For unknown Hurst parameter, the likelihood does not lend itself amenable to the popular Girsanov form, rendering usual asymptotic development difficult. As such, we develop increasing domain infill asymptotic theory, by discretizing the SDE. In this setup, we establish consistency and asymptotic normality of the maximum likelihood estimators, as well as consistency and asymptotic normality of the Bayesian posterior distributions. However, classical or Bayesian asymptotic normality with respect to the Hurst parameter could not be established. We supplement our theoretical investigations with simulation studies in a non-asymptotic setup, prescribing suitable methodologies for classical and Bayesian analyses of SDEs driven by fractional Brownian motion. Applications to a real, close price data, along with comparison with standard SDE driven by Wiener process, is also considered. As expected, it turned out that our Bayesian fractional SDE triumphed over the other model and methods, in both simulated and real data applications.

math.ST↗

arXiv2022

Additive Security Games: Structure and Optimization

In this work, we provide a structural characterization of the possible Nash equilibria in the well-studied class of security games with additive utility. Our analysis yields a classification of possible equilibria into seven types and we provide closed-form feasibility conditions for each type as well as closed-form expressions for the expected outcomes to the players at equilibrium. We provide uniqueness and multiplicity results for each type and utilize our structural approach to propose a novel algorithm to compute equilibria of each type when they exist. We then consider the special cases of security games with fully protective resources and zero-sum games. Under the assumption that the defender can perturb the payoffs to the attacker, we study the problem of optimizing the defender expected outcome at equilibrium. We show that this problem is weakly NP- hard in the case of Stackelberg equilibria and multiple attacker resources and present a pseudopolynomial time procedure to solve this problem for the case of Nash equilibria under mild assumptions. Finally, to address non-additive security games, we propose a notion of nearest additive game and demonstrate the existence and uniqueness of a such a nearest additive game for any non-additive game.

cs.GT↗

arXiv2022

IID Sampling from Posterior Dirichlet Process Mixtures

The influence of Dirichlet process mixture is ubiquitous in the Bayesian nonparametrics literature. But sampling from its posterior distribution remains a challenge, despite the advent of various Markov chain Monte Carlo methods. The primary challenge is the infinite-dimensional setup, and even if the infinite-dimensional random measure is integrated out, high-dimensionality and discreteness still remain difficult issues to deal with. In this article, exploiting the key ideas proposed in Bhattacharya (2021b), we propose a novel methodology for drawing iid realizations from posteriors of Dirichlet process mixtures. We focus in particular on the more general and flexible model of Bhattacharya (2008), so that the methods developed here are simply applicable to the traditional Dirichlet process mixture. We illustrate our ideas on the well-known enzyme, acidity and the galaxy datasets, which are usually considered benchmark datasets for mixture applications. Generating 10, 000 iid realizations from the Dirichlet process mixture posterior of Bhattacharya (2008) given these datasets took 19 minutes, 8 minutes and 5 minutes, respectively, in our parallel implementation.

stat.CO↗

arXiv2021

Hierarchical Bayesian state-space modeling of age- and sex-structured wildlife population dynamics

Biodiversity is declining at alarming rates worldwide, including for large wild mammals. It is therefore imperative to develop effective population conservation and recovery strategies. Population dynamics models can provide insights into processes driving declines of particular populations of a species and their relative importance. We develop an integrated Bayesian state-space population dynamics model for wildlife populations and illustrate it using a topi population inhabiting the Masai Mara Ecosystem in Kenya. The model is general and integrates ground demographic survey with aerial survey monitoring data. It incorporates population age- and sex-structure and life-history traits and relates birth rates, age-specific survival rates and sex ratio with meteorological covariates, prior population density, environmental seasonality and predation risk. The model runs on a monthly time step, enabling accurate characterization of reproductive seasonality, phenology, synchrony and prolificacy of births and juvenile recruitment. Model performance is evaluated using balanced bootstrap sampling and comparing predictions with aerial population size estimates. The model is implemented using MCMC methods and reproduces several well-known features of the Mara topi population, including striking and persistent population decline, seasonality of births and juvenile recruitment. It can be readily adapted for other wildlife species and extended to incorporate several additional useful features.

stat.AP↗

arXiv2021

IID Sampling from Intractable Multimodal and Variable-Dimensional Distributions

Bhattacharya (2021b) has introduced a novel methodology for generating iid realizations from any target distribution on the Euclidean space, irrespective of dimensionality. In this article, our purpose is two-fold. We first extend the method for obtaining iid realizations from general multimodal distributions, and illustrate with a mixture of two 50-dimensional normal distributions. Then we extend the iid sampling method for fixed-dimensional distributions to variable-dimensional situations and illustrate with a variable-dimensional normal mixture modeling of the well-known "acidity data", with further demonstration of the applicability of the iid sampling method developed for multimodal distributions.

stat.ME↗

arXiv2021

IID Sampling from Doubly Intractable Distributions

Intractable posterior distributions of parameters with intractable normalizing constants depending upon the parameters are known as doubly intractable posterior distributions. The terminology itself indicates that obtaining Bayesian inference from such posteriors is doubly difficult compared to traditional intractable posteriors where the normalizing constants are tractable and admit traditional Markov Chain Monte Carlo (MCMC) solutions. As can be anticipated, a plethora of MCMC-based methods have originated in the literature to deal with doubly intractable distributions. Yet, it remains very much unclear if any of the methods can satisfactorily sample from such posteriors, particularly in high-dimensional setups. In this article, we consider efficient Monte Carlo and importance sampling approximations of the intractable normalizing constant for a few values of the parameters, and Gaussian process interpolations for the remaining values of the parameters, using the approximations. We then incorporate this strategy within the exact iid sampling framework developed in Bhattacharya (2021a) and Bhattacharya (2021b), and illustrate the methodology with simulation experiments comprising a two-dimensional normal-gamma posterior, a two-dimensional Ising model posterior, a two-dimensional Strauss process posterior and a 100-dimensional autologistic model posterior. In each case we demonstrate great accuracy of our methodology, which is also computationally extremely efficient, often taking only a few minutes for generating 10, 000 iid realizations on 80 processors.

stat.CO↗