arXiv subjects

Tapio Helin

Publications and source records attributed to Tapio Helin.

At least 19 recordsLinked to original sources

arXiv2026

Statistical inverse learning and $\ell^1$-regularization

We study the recovery of sparse functions from finite, noisy, and indirect observations in the framework of statistical inverse learning. The unknown is modeled as an element of $\ell^1$, and observations are generated through a possibly nonlinear forward operator $A:\ell^1\to H$, where $H$ is a vector-valued reproducing kernel Hilbert space. We propose an $\ell^1$-regularized empirical risk minimizer and develop a theoretical analysis of its statistical properties. Under mild assumptions, we establish almost-sure consistency and derive non-asymptotic high-probability convergence rates in both the prediction and $\ell^1$ reconstruction norms. The rates depend on the source smoothness parameter $r$, characterized by a variational source condition, and the effective dimension exponent $b$, describing the polynomial spectral decay of the covariance operator. We further prove matching minimax lower bounds, showing that the obtained convergence rates are optimal. To relate the theory to practical sparsity models, we consider finitely smoothing operators of the form $A=G\circ S$, where $S$ is a synthesis operator, and show that approximation-space assumptions imply the required variational source conditions. In particular, we prove that membership in the approximation space $k_t$ is equivalent to polynomial decay of the best $n$-term approximation error. Finally, we verify the assumptions for two representative inverse problems: reaction coefficient identification in elliptic PDEs and sparse computed tomography. For filtered Radon transforms, we derive explicit effective-dimension asymptotics, yielding concrete convergence rates for standard image models and sparsifying systems.

stat.ML↗

arXiv2026

A Convex Approximation Framework for Neural Likelihood-Based Bayesian Inverse Problems

Many problems in science and engineering are difficult to model accurately, either due to unknown physical mechanisms, poorly quantified measurement uncertainty, or prohibitive computational costs of high-fidelity simulations. These challenges limit the applicability of classical probabilistic inference methods such as Markov chain Monte Carlo, especially in high-dimensional Bayesian inverse problems. As data from scientific experiments become increasingly available, machine learning methods offer a flexible alternative to explicit parametric modelling. We study neural likelihood approximation, where the goal is to learn the likelihood function directly from data without explicit knowledge of the underlying data-generating process. A common approach trains likelihood surrogates by minimizing the Kullback-Leibler divergence between the true posterior and an approximate posterior, which is equivalent to minimizing the expected negative log-likelihood. This work improves the theoretical foundations of neural likelihood approximation by alleviating limitations of restrictive model classes: we show that, by working with un-normalized potentials and folding normalization into the training objective, the resulting learning problem is strictly convex. We show that empirical minimizers of the resulting data-driven objective converge to the true likelihood as the sample size grows. Numerical experiments for the neural likelihood approximation are conducted for a deblurring and a non-linear PDE based imaging problem.

stat.ML↗

arXiv2026

Robust Model-Based Iteration for Passive Gamma Emission Tomography

Passive Gamma Emission Tomography (PGET) is an IAEA-approved technique for verifying spent nuclear fuel assemblies prior to geological disposal. Reconstructing the emission and attenuation maps from PGET measurements is a nonlinear ill-posed inverse problem, currently solved with a Levenberg-Marquardt (LM) scheme that requires 10-20 iterations to achieve sufficient accuracy. We propose an accelerated iterative solver that combines the LM algorithm with a Deep Gauss-Newton step, in which a learned operator refines the update proposed by the deterministic algorithm at each iteration. A safeguard condition based on the trust-region model ensures that the accelerated iterates perform no worse than LM and retain convergence to a critical point of the regularized objective. Within this framework we compare three architectures for the learned component: an encoder-decoder-style convolutional neural network, Fourier Neural Operators, and Wavelet Neural Operators. Each is trained on a small set of coarsely simulated 9x9 assemblies. Experiments on simulated and real measurements from Finnish nuclear power plants show that the proposed scheme reaches LM-quality reconstructions in roughly one third of the iterations, while revealing architecture-dependent trade-offs in robustness against out-of-distribution inputs.

math.NA↗

arXiv2026

Batch-based Bayesian Optimal Experimental Design in Linear Inverse Problems

Experimental design is central to science and engineering. A ubiquitous challenge is how to maximize the value of information obtained from expensive or constrained experimental settings. Bayesian optimal experimental design (OED) provides a principled framework for addressing such questions. In this paper, we study experimental design problems such as the optimization of sensor locations over a continuous domain in the context of linear Bayesian inverse problems. We focus in particular on batch design, that is, the simultaneous optimization of multiple design variables, which leads to a notoriously difficult non-convex optimization problem. We tackle this challenge using a promising strategy recently proposed in the frequentist setting, which relaxes A-optimal design to the space of finite positive measures. Our main contribution is the rigorous identification of the Bayesian inference problem corresponding to this relaxed A-optimal OED formulation. Moreover, building on recent work, we develop a Wasserstein gradient-flow -based optimization algorithm for the expected utility and introduce novel regularization schemes that guarantee convergence to an empirical measure. These theoretical results are supported by numerical experiments demonstrating both convergence and the effectiveness of the proposed regularization strategy.

stat.ME↗

arXiv2026

Score-based diffusion models for diffuse optical tomography with uncertainty quantification

Score-based diffusion models are a recently developed framework for posterior sampling in Bayesian inverse problems with a state-of-the-art performance for severely ill-posed problems by leveraging a powerful prior distribution learned from empirical data. Despite generating significant interest especially in the machine-learning community, a thorough study of realistic inverse problems in the presence of modelling error and utilization of physical measurement data is still outstanding. In this work, the framework of unconditional representation for the conditional score function (UCoS) is evaluated for linearized difference imaging in diffuse optical tomography (DOT). DOT uses boundary measurements of near-infrared light to estimate the spatial distribution of absorption and scattering parameters in biological tissues. The problem is highly ill-posed and thus sensitive to noise and modelling errors. We introduce a novel regularization approach that prevents overfitting of the score function by constructing a mixed score composed of a learned and a model-based component. Validation of this approach is done using both simulated and experimental measurement data. The experiments demonstrate that a data-driven prior distribution results in posterior samples with low variance, compared to classical model-based estimation, and centred around the ground truth, even in the context of a highly ill-posed problem and in the presence of modelling errors.

stat.ML↗

arXiv2025

Approximation of differential entropy in Bayesian optimal experimental design

Bayesian optimal experimental design provides a principled framework for selecting experimental settings that maximize obtained information. In this work, we focus on estimating the expected information gain in the setting where the differential entropy of the likelihood is either independent of the design or can be evaluated explicitly. This reduces the problem to maximum entropy estimation, alleviating several challenges inherent in expected information gain computation. Our study is motivated by large-scale inference problems, such as inverse problems, where the computational cost is dominated by expensive likelihood evaluations. We propose a computational approach in which the evidence density is approximated by a Monte Carlo or quasi-Monte Carlo surrogate, while the differential entropy is evaluated using standard methods without additional likelihood evaluations. We prove that this strategy achieves convergence rates that are comparable to, or better than, state-of-the-art methods for full expected information gain estimation, particularly when the cost of entropy evaluation is negligible. Moreover, our approach relies only on mild smoothness of the forward map and avoids stronger technical assumptions required in earlier work. We also present numerical experiments, which confirm our theoretical findings.

stat.ML↗

arXiv2025

Bayesian optimal experimental design with Wasserstein information criteria

Bayesian optimal experimental design (OED) provides a principled framework for selecting observations or experiments. We introduce new Bayesian design criteria based on the expected Wasserstein-$p$ distance between the prior and posterior distributions, termed Wasserstein information criteria. These criteria have many parallels with the widely used expected information gain (EIG) criterion, which instead relies on the Kullback--Leibler divergence. We show that the Wasserstein-$2$ criterion admits a closed-form solution in the linear-Gaussian setting, a property which can be used for more general approximation schemes, and contrast this solution with classical notions of Bayesian alphabetic optimality. Then we develop a stability analysis of the Wasserstein-$1$ criterion, wherein we bound errors induced by perturbations of the prior or likelihood. We partially extend this analysis to the Wasserstein-$2$ criterion. In particular, these results yield error rates for empirical approximations of the prior. We then illustrate the computability of the Wasserstein-$2$ criterion and demonstrate our approximation rates through simulations.

stat.ME↗

arXiv2025

Recovering a (1+1)-dimensional wave equation from a single white noise boundary measurement

We consider the following inverse problem: Suppose a $(1+1)$-dimensional wave equation on $\mathbb{R}_+$ with zero initial conditions is excited with a Neumann boundary data modelled as a white noise process. Given also the Dirichlet data at the same point, determine the unknown first order coefficient function of the system. We first establish that direct problem is well-posed. The inverse problem is then solved by showing that correlations of the boundary data determine the Neumann-to-Dirichlet operator in the sense of distributions, which is known to uniquely identify the coefficient. This approach has applications in acoustic measurements of internal cross-sections of fluid pipes such as pressurised water supply pipes and vocal tract shape determination.

math.AP↗

arXiv2024

Gradient-Based Non-Linear Inverse Learning

We study statistical inverse learning in the context of nonlinear inverse problems under random design. Specifically, we address a class of nonlinear problems by employing gradient descent (GD) and stochastic gradient descent (SGD) with mini-batching, both using constant step sizes. Our analysis derives convergence rates for both algorithms under classical a priori assumptions on the smoothness of the target function. These assumptions are expressed in terms of the integral operator associated with the tangent kernel, as well as through a bound on the effective dimension. Additionally, we establish stopping times that yield minimax-optimal convergence rates within the classical reproducing kernel Hilbert space (RKHS) framework. These results demonstrate the efficacy of GD and SGD in achieving optimal rates for nonlinear inverse problems in random design.

stat.ML↗

arXiv2024

Learning sparsity-promoting regularizers for linear inverse problems

This paper introduces a novel approach to learning sparsity-promoting regularizers for solving linear inverse problems. We develop a bilevel optimization framework to select an optimal synthesis operator, denoted as $B$, which regularizes the inverse problem while promoting sparsity in the solution. The method leverages statistical properties of the underlying data and incorporates prior knowledge through the choice of $B$. We establish the well-posedness of the optimization problem, provide theoretical guarantees for the learning process, and present sample complexity bounds. The approach is demonstrated through theoretical infinite-dimensional examples, including compact perturbations of a known operator and the problem of learning the mother wavelet, and through extensive numerical simulations. This work extends previous efforts in Tikhonov regularization by addressing non-differentiable norms and proposing a data-driven approach for sparse regularization in infinite dimensions.

stat.ML↗

arXiv2024

The GPU-based High-order adaptive OpticS Testbench

The GPU-based High-order adaptive OpticS Testbench (GHOST) at the European Southern Observatory (ESO) is a new 2-stage extreme adaptive optics (XAO) testbench at ESO. The GHOST is designed to investigate and evaluate new control methods (machine learning, predictive control) for XAO which will be required for instruments such as the Planetary Camera and Spectrograph of ESOs Extremely Large Telescope. The first stage corrections are performed in simulation, with the residual wavefront error at each iteration saved. The residual wavefront errors from the first stage are then injected into the GHOST using a spatial light modulator. The second stage correction is made with a Boston Michromachines Corporation 492 actuator deformable mirror and a pyramid wavefront sensor. The flexibility of the bench also opens it up to other applications, one such application is investigating the flip-flop modulation method for the pyramid wavefront sensor.

astro-ph.IM↗

arXiv2024

The MICADO first light imager for the ELT: off-axis performance of PSF reconstruction

The highest scientific return, for adaptive optics (AO) observations, is achieved with a reliable reconstruction of the PSF. This is especially true for MICADO@ELT. In this presentation, we will focus on extending the MICADO PSF reconstruction (PSF-R) method to the off-axis case. Specifically, a novel approach based on temporal-based tomography of AO telemetry data has been recently implemented. Results from the PSF-R of both simulated and real data show that, at half isoplanatic angle distances, a precision of about 10-15% is achievable in both Strehl ratio and full-width at half maximum, paving the way to extend the MICADO PSF-R tool also to the multi-conjugated AO case.

astro-ph.IM↗

arXiv2024

Surrogate model for Bayesian optimal experimental design in chromatography

We applied Bayesian Optimal Experimental Design (OED) in the estimation of parameters involved in the Equilibrium Dispersive Model for chromatography with two components with the Langmuir adsorption isotherm. The coefficients estimated were Henry's coefficients, the total absorption capacity and the number of theoretical plates, while the design variables were the injection time and the initial concentration. The Bayesian OED algorithm is based on nested Monte Carlo estimation, which becomes computationally challenging due to the simulation time of the PDE involved in the dispersive model. This complication was relaxed by introducing a surrogate model based on Piecewise Sparse Linear Interpolation. Using the surrogate model instead the original reduces significantly the simulation time and it approximates the solution of the PDE with high degree of accuracy. The estimation of the parameters over strategical design points provided by OED reduces the uncertainty in the estimation of parameters. Additionally, the Bayesian OED methodology indicates no improvements when increasing the number of measurements in temporal nodes above a threshold value.

stat.AP↗

arXiv2024

The power of prediction: spatiotemporal Gaussian process modeling for predictive control in slope-based wavefront sensing

Time-delay error is a significant error source in adaptive optics (AO) systems. It arises from the latency between sensing the wavefront and applying the correction. Predictive control algorithms reduce the time-delay error, providing significant performance gains, especially for high-contrast imaging. However, the predictive controller's performance depends on factors such as the WFS type, the measurement noise, the AO system's geometry, and the atmospheric conditions. This work studies the limits of prediction under different imaging conditions through spatiotemporal Gaussian process models. The method provides a predictive reconstructor that is optimal in the least-squares sense, conditioned on the fixed times series of WFS data and our knowledge of the atmosphere. We demonstrate that knowledge is power in predictive AO control. With an SHS-based extreme AO instrument, perfect knowledge of Frozen Flow evolution (wind and Cn2 profile) leads to a reduction of the residual wavefront phase variance up to a factor of 3.5 compared to a non-predictive approach. If there is uncertainty in the profile or evolution models, the gain is more modest. Still, assuming that only effective wind speed is available (without direction) led to reductions in variance by a factor of 2.3. We also study the value of data for predictive filters by computing the experimental utility for different scenarios to answer questions such as: How many past data frames should the prediction filter consider, and is it always most advantageous to use the most recent data? We show that within the scenarios considered, more data consistently increases prediction accuracy. Further, we demonstrate that given a computational limitation on how many past frames we can use, an optimized selection of $n$ past frames leads to a 10-15% additional improvement in RMS over using the n latest consecutive frames of data.

astro-ph.IM↗

arXiv2024

An Unconditional Representation of the Conditional Score in Infinite-Dimensional Linear Inverse Problems

Score-based diffusion models (SDMs) have emerged as a powerful tool for sampling from the posterior distribution in Bayesian inverse problems. However, existing methods often require multiple evaluations of the forward mapping to generate a single sample, resulting in significant computational costs for large-scale inverse problems. To address this, we propose an unconditional representation of the conditional score function (UCoS) tailored to linear inverse problems, which avoids forward model evaluations during sampling by shifting computational effort to an offline training phase. In this phase, a \emph{task-dependent} score function is learned based on the linear forward operator. Crucially, we show that the conditional score can be derived \emph{exactly} from a trained (unconditional) score using affine transformations, eliminating the need for conditional score approximations. Our approach is formulated in infinite-dimensional function spaces, making it inherently discretization-invariant. We support this formulation with a rigorous convergence analysis that justifies UCoS beyond any specific discretization. Finally we validate UCoS through high-dimensional computed tomography (CT) and image deblurring experiments, demonstrating both scalability and accuracy.

stat.ML↗

arXiv2024

Recovery of transversely-isotropic elastic material parameters in induction motor rotors

We propose numerical algorithms for recovering parameters in eigenvalue problems for linear elasticity of transversely isotropic materials. Specifically, the algorithms are used to recover the elastic constants of a rotor core. Numerical tests show that in the noiseless setup, two pairs of bending modes are sufficient for recovering one to four parameters accurately. To recover all five parameters that govern the elastic properties of electric engines accurately, we require three pairs of bending modes and one torsional mode. Moreover, we study the stability of the inversion method against multiplicative noise; for tests in which the data contained multiplicative noise of at most $1\%$, we find that all parameters can be recovered with an error less than $10\%$.

math.NA↗

arXiv2023

Laboratory Experiments of Model-based Reinforcement Learning for Adaptive Optics Control

Direct imaging of Earth-like exoplanets is one of the most prominent scientific drivers of the next generation of ground-based telescopes. Typically, Earth-like exoplanets are located at small angular separations from their host stars, making their detection difficult. Consequently, the adaptive optics (AO) system's control algorithm must be carefully designed to distinguish the exoplanet from the residual light produced by the host star. A new promising avenue of research to improve AO control builds on data-driven control methods such as Reinforcement Learning (RL). RL is an active branch of the machine learning research field, where control of a system is learned through interaction with the environment. Thus, RL can be seen as an automated approach to AO control, where its usage is entirely a turnkey operation. In particular, model-based reinforcement learning (MBRL) has been shown to cope with both temporal and misregistration errors. Similarly, it has been demonstrated to adapt to non-linear wavefront sensing while being efficient in training and execution. In this work, we implement and adapt an RL method called Policy Optimization for AO (PO4AO) to the GHOST test bench at ESO headquarters, where we demonstrate a strong performance of the method in a laboratory environment. Our implementation allows the training to be performed parallel to inference, which is crucial for on-sky operation. In particular, we study the predictive and self-calibrating aspects of the method. The new implementation on GHOST running PyTorch introduces only around 700 microseconds in addition to hardware, pipeline, and Python interface latency. We open-source well-documented code for the implementation and specify the requirements for the RTC pipeline. We also discuss the important hyperparameters of the method, the source of the latency, and the possible paths for a lower latency implementation.

astro-ph.IM↗

arXiv2023

Statistical inverse learning problems with random observations

We provide an overview of recent progress in statistical inverse problems with random experimental design, covering both linear and nonlinear inverse problems. Different regularization schemes have been studied to produce robust and stable solutions. We discuss recent results in spectral regularization methods and regularization by projection, exploring both approaches within the context of Hilbert scales and presenting new insights particularly in regularization by projection. Additionally, we overview recent advancements in regularization using convex penalties. Convergence rates are analyzed in terms of the sample size in a probabilistic sense, yielding minimax rates in both expectation and probability. To achieve these results, the structure of reproducing kernel Hilbert spaces is leveraged to establish minimax rates in the statistical learning setting. We detail the assumptions underpinning these key elements of our proofs. Finally, we demonstrate the application of these concepts to nonlinear inverse problems in pharmacokinetic/pharmacodynamic (PK/PD) models, where the task is to predict changes in drug concentrations in patients.

math.ST↗