PyTorchKR Papers

최신 논문

총 1,831편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2405.13846

Regression Trees Know Calculus

Regression trees have emerged as a preeminent tool for solving real-world regression problems due to their ability to deal with nonlinearities, interaction effects and sharp discontinuities. In this article, we rather study regression trees applied to well-behaved, differentiable functions, and determine the relationship between node parameters and the local gradient of the function being approximated. We find a simple estimate of the gradient which can be efficiently computed using quantities exposed by popular tree learning libraries. This allows the tools developed in the context of differentiable algorithms, like neural nets and Gaussian processes, to be deployed to tree-based models. To demonstrate this, we study measures of model sensitivity defined in terms of integrals of gradients and demonstrate how to compute them for regression trees using the proposed gradient estimates. Quantitative and qualitative numerical experiments reveal the capability of gradients estimated by regression trees to improve predictive analysis, solve tasks in uncertainty quantification, and provide interpretation of model behavior.

arXiv:2405.14492공식 코드 · Papers with Code

Iterative Methods for Full-Scale Gaussian Process Approximations for Large Spatial Data

Gaussian processes are flexible probabilistic regression models which are widely used in statistics and machine learning. However, a drawback is their limited scalability to large data sets. To alleviate this, full-scale approximations (FSAs) combine predictive process methods and covariance tapering, thus approximating both global and local structures. We show how iterative methods can be used to reduce computational costs in calculating likelihoods, gradients, and predictive distributions with FSAs. In particular, we introduce a novel preconditioner and show theoretically and empirically that it accelerates the conjugate gradient method's convergence speed and mitigates its sensitivity with respect to the FSA parameters and the eigenvalue structure of the original covariance matrix, and we demonstrate empirically that it outperforms a state-of-the-art pivoted Cholesky preconditioner. Furthermore, we introduce an accurate and fast way to calculate predictive variances using stochastic simulation and iterative methods. In addition, we show how our newly proposed FITC preconditioner can also be used in iterative methods for Vecchia approximations. In our experiments, it outperforms existing state-of-the-art preconditioners for Vecchia approximations. All methods are implemented in a free C++ software library with high-level Python and R packages.

arXiv:2405.14250

Diffusion models for Gaussian distributions: Exact solutions and Wasserstein errors

Diffusion or score-based models recently showed high performance in image generation. They rely on a forward and a backward stochastic differential equations (SDE). The sampling of a data distribution is achieved by numerically solving the backward SDE or its associated flow ODE. Studying the convergence of these models necessitates to control four different types of error: the initialization error, the truncation error, the discretization error and the score approximation. In this paper, we theoretically study the behavior of diffusion models and their numerical implementation when the data distribution is Gaussian. Our first contribution is to derive the analytical solutions of the backward SDE and the probability flow ODE and to prove that these solutions and their discretizations are all Gaussian processes. Our second contribution is to compute the exact Wasserstein errors between the target and the numerically sampled distributions for any numerical scheme. This allows us to monitor convergence directly in the data space, while experimental works limit their empirical analysis to Inception features. An implementation of our code is available online.

arXiv:2405.12614

Efficient modeling of sub-kilometer surface wind with Gaussian processes and neural networks

Accurately representing surface weather at the sub-kilometer scale is crucial for optimal decision-making in a wide range of applications. This motivates the use of statistical techniques to provide accurate and calibrated probabilistic predictions at a lower cost compared to numerical simulations. Wind represents a particularly challenging variable to model due to its high spatial and temporal variability. This paper presents a novel approach that integrates Gaussian processes and neural networks to model surface wind gusts at sub-kilometer resolution, leveraging multiple data sources, including numerical weather prediction models, topographical descriptors, and in-situ measurements. Results demonstrate the added value of modeling the multivariate covariance structure of the variable of interest, as opposed to only applying a univariate probabilistic regression approach. Modeling the covariance enables the optimal integration of observed measurements from ground stations, which is shown to reduce the continuous ranked probability score compared to the baseline. Moreover, it allows the generation of realistic fields that are also marginally calibrated, aided by scalable techniques such as random Fourier features and pathwise conditioning. We discuss the effect of different modeling choices, as well as different degrees of approximation, and present our results for a case study.

arXiv:2405.11975

Optimal Privacy-Aware Stochastic Sampling

This paper presents a stochastic sampling framework for privacy-aware data sharing, where a sensor observes a process correlated with private information. A sampler determines whether to retain or discard sensor observations, balancing the tradeoff between data utility and privacy. Retained samples are shared with an adversary who may attempt to infer the private process, with privacy leakage quantified using mutual information. The sampler design is formulated as an optimization problem with two objectives: $\left(\romannumeral1\right)$ minimizing the reconstruction error of the observed process using the sampler's output, $\left(\romannumeral2\right)$ reducing the privacy leakages. For a general class of processes, we show that the optimal reconstruction policy is deterministic and derive the optimality conditions for the sampling policy using a dynamic decomposition method, which enables the sampler to control the adversary's belief about private inputs. For linear Gaussian processes, we propose a simplified design by restricting the sampling policy to a specific collection, providing analytical expressions for the reconstruction error, belief state, and sampling objectives based on conditional means and covariances. Additionally, we develop a numerical optimization algorithm to optimize the sampling and reconstruction policies, wherein the policy gradient theorem for the optimal sampling design is derived based on the implicit function theorem. Simulations demonstrate the effectiveness of the proposed method in achieving accurate state reconstruction, privacy protection, and data size reduction.

arXiv:2405.12802

Stochastic Inference of Plate Bending from Heterogeneous Data: Physics-informed Gaussian Processes via Kirchhoff-Love Theory

Advancements in machine learning and an abundance of structural monitoring data have inspired the integration of mechanical models with probabilistic models to identify a structure's state and quantify the uncertainty of its physical parameters and response. In this paper, we propose an inference methodology for classical Kirchhoff-Love plates via physics-informed Gaussian Processes (GP). A probabilistic model is formulated as a multi-output GP by placing a GP prior on the deflection and deriving the covariance function using the linear differential operators of the plate governing equations. The posteriors of the flexural rigidity, hyperparameters, and plate response are inferred in a Bayesian manner using Markov chain Monte Carlo (MCMC) sampling from noisy measurements. We demonstrate the applicability with two examples: a simply supported plate subjected to a sinusoidal load and a fixed plate subjected to a uniform load. The results illustrate how the proposed methodology can be employed to perform stochastic inference for plate rigidity and physical quantities by integrating measurements from various sensor types and qualities. Potential applications of the presented methodology are in structural health monitoring and uncertainty quantification of plate-like structures.

arXiv:2405.11683

Conditionally-Conjugate Gaussian Process Factor Analysis for Spike Count Data via Data Augmentation

Gaussian process factor analysis (GPFA) is a latent variable modeling technique commonly used to identify smooth, low-dimensional latent trajectories underlying high-dimensional neural recordings. Specifically, researchers model spiking rates as Gaussian observations, resulting in tractable inference. Recently, GPFA has been extended to model spike count data. However, due to the non-conjugacy of the likelihood, the inference becomes intractable. Prior works rely on either black-box inference techniques, numerical integration or polynomial approximations of the likelihood to handle intractability. To overcome this challenge, we propose a conditionally-conjugate Gaussian process factor analysis (ccGPFA) resulting in both analytically and computationally tractable inference for modeling neural activity from spike count data. In particular, we develop a novel data augmentation based method that renders the model conditionally conjugate. Consequently, our model enjoys the advantage of simple closed-form updates using a variational EM algorithm. Furthermore, due to its conditional conjugacy, we show our model can be readily scaled using sparse Gaussian Processes and accelerated inference via natural gradients. To validate our method, we empirically demonstrate its efficacy through experiments.

arXiv:2405.10581

Future Aware Safe Active Learning of Time Varying Systems using Gaussian Processes

Experimental exploration of high-cost systems with safety constraints, common in engineering applications, is a challenging endeavor. Data-driven models offer a promising solution, but acquiring the requisite data remains expensive and is potentially unsafe. Safe active learning techniques prove essential, enabling the learning of high-quality models with minimal expensive data points and high safety. This paper introduces a safe active learning framework tailored for time-varying systems, addressing drift, seasonal changes, and complexities due to dynamic behavior. The proposed Time-aware Integrated Mean Squared Prediction Error (T-IMSPE) method minimizes posterior variance over current and future states, optimizing information gathering also in the time domain. Empirical results highlight T-IMSPE's advantages in model quality through toy and real-world examples. State of the art Gaussian processes are compatible with T-IMSPE. Our theoretical contributions include a clear delineation which Gaussian process kernels, domains, and weighting measures are suitable for T-IMSPE and even beyond for its non-time aware predecessor IMSPE.

arXiv:2405.09989

A Gaussian Process Model for Ordinal Data with Applications to Chemoinformatics

With the proliferation of screening tools for chemical testing, it is now possible to create vast databases of chemicals easily. However, rigorous statistical methodologies employed to analyse these databases are in their infancy, and further development to facilitate chemical discovery is imperative. In this paper, we present conditional Gaussian process models to predict ordinal outcomes from chemical experiments, where the inputs are chemical compounds. We implement the Tanimoto distance, a metric on the chemical space, within the covariance of the Gaussian processes to capture correlated effects in the chemical space. A novel aspect of our model is that the kernel contains a scaling parameter, a feature not previously examined in the literature, that controls the strength of the correlation between elements of the chemical space. Using molecular fingerprints, a numerical representation of a compound's location within the chemical space, we find that accounting for correlation amongst chemical compounds improves predictive performance over the uncorrelated model, where effects are assumed to be independent. Moreover, we present a genetic algorithm for the facilitation of chemical discovery and identification of important features to the compound's efficacy, based on two criteria derived from the proposed model. Simulation studies are conducted to demonstrate the suitability of the proposed methods. Our model is demonstrated on a hazard classification problem of organic solvents.

arXiv:2405.10264

Architectures and random properties of symplectic quantum circuits

Parametrized and random unitary (or orthogonal) $n$-qubit circuits play a central role in quantum information. As such, one could naturally assume that circuits implementing symplectic transformation would attract similar attention. However, this is not the case, as $\mathbb{SP}(d/2)$ -- the group of $d\times d$ unitary symplectic matrices -- has thus far been overlooked. In this work, we aim at starting to right this wrong. We begin by presenting a universal set of generators $\mathcal{G}$ for the symplectic algebra $i\mathfrak{sp}(d/2)$, consisting of one- and two-qubit Pauli operators acting on neighboring sites in a one-dimensional lattice. Here, we uncover two critical differences between such set, and equivalent ones for unitary and orthogonal circuits. Namely, we find that the operators in $\mathcal{G}$ cannot generate arbitrary local symplectic unitaries and that they are not translationally invariant. We then review the Schur-Weyl duality between the symplectic group and the Brauer algebra, and use tools from Weingarten calculus to prove that Pauli measurements at the output of Haar random symplectic circuits can converge to Gaussian processes. As a by-product, such analysis provides us with concentration bounds for Pauli measurements in circuits that form $t$-designs over $\mathbb{SP}(d/2)$. To finish, we present tensor-network tools to analyze shallow random symplectic circuits, and we use these to numerically show that computational-basis measurements anti-concentrate at logarithmic depth.

arXiv:2405.10229

Random ReLU Neural Networks as Non-Gaussian Processes

We consider a large class of shallow neural networks with randomly initialized parameters and rectified linear unit activation functions. We prove that these random neural networks are well-defined non-Gaussian processes. As a by-product, we demonstrate that these networks are solutions to stochastic differential equations driven by impulsive white noise (combinations of random Dirac measures). These processes are parameterized by the law of the weights and biases as well as the density of activation thresholds in each bounded region of the input domain. We prove that these processes are isotropic and wide-sense self-similar with Hurst exponent 3/2. We also derive a remarkably simple closed-form expression for their autocovariance function. Our results are fundamentally different from prior work in that we consider a non-asymptotic viewpoint: The number of neurons in each bounded region of the input domain (i.e., the width) is itself a random variable with a Poisson law with mean proportional to the density parameter. Finally, we show that, under suitable hypotheses, as the expected width tends to infinity, these processes can converge in law not only to Gaussian processes, but also to non-Gaussian processes depending on the law of the weights. Our asymptotic results provide a new take on several classical results (wide networks converge to Gaussian processes) as well as some new ones (wide networks can converge to non-Gaussian processes).

arXiv:2405.09541

Spectral complexity of deep neural networks

It is well-known that randomly initialized, push-forward, fully-connected neural networks weakly converge to isotropic Gaussian processes, in the limit where the width of all layers goes to infinity. In this paper, we propose to use the angular power spectrum of the limiting field to characterize the complexity of the network architecture. In particular, we define sequences of random variables associated with the angular power spectrum, and provide a full characterization of the network complexity in terms of the asymptotic distribution of these sequences as the depth diverges. On this basis, we classify neural networks as low-disorder, sparse, or high-disorder; we show how this classification highlights a number of distinct features for standard activation functions, and in particular, sparsity properties of ReLU networks. Our theoretical results are also validated by numerical simulations.

arXiv:2405.09109

Motion Prediction with Gaussian Processes for Safe Human-Robot Interaction in Virtual Environments

Humans use collaborative robots as tools for accomplishing various tasks. The interaction between humans and robots happens in tight shared workspaces. However, these machines must be safe to operate alongside humans to minimize the risk of accidental collisions. Ensuring safety imposes many constraints, such as reduced torque and velocity limits during operation, thus increasing the time to accomplish many tasks. However, for applications such as using collaborative robots as haptic interfaces with intermittent contacts for virtual reality applications, speed limitations result in poor user experiences. This research aims to improve the efficiency of a collaborative robot while improving the safety of the human user. We used Gaussian process models to predict human hand motion and developed strategies for human intention detection based on hand motion and gaze to improve the time for the robot and human security in a virtual environment. We then studied the effect of prediction. Results from comparisons show that the prediction models improved the robot time by 3\% and safety by 17\%. When used alongside gaze, prediction with Gaussian process models resulted in an improvement of the robot time by 2\% and the safety by 13\%.

arXiv:2405.08711

Data-driven Force Observer for Human-Robot Interaction with Series Elastic Actuators using Gaussian Processes

Ensuring safety and adapting to the user's behavior are of paramount importance in physical human-robot interaction. Thus, incorporating elastic actuators in the robot's mechanical design has become popular, since it offers intrinsic compliance and additionally provide a coarse estimate for the interaction force by measuring the deformation of the elastic components. While observer-based methods have been shown to improve these estimates, they rely on accurate models of the system, which are challenging to obtain in complex operating environments. In this work, we overcome this issue by learning the unknown dynamics components using Gaussian process (GP) regression. By employing the learned model in a Bayesian filtering framework, we improve the estimation accuracy and additionally obtain an observer that explicitly considers local model uncertainty in the confidence measure of the state estimate. Furthermore, we derive guaranteed estimation error bounds, thus, facilitating the use in safety-critical applications. We demonstrate the effectiveness of the proposed approach experimentally in a human-exoskeleton interaction scenario.

arXiv:2405.08318

No-Regret Learning of Nash Equilibrium for Black-Box Games via Gaussian Processes

This paper investigates the challenge of learning in black-box games, where the underlying utility function is unknown to any of the agents. While there is an extensive body of literature on the theoretical analysis of algorithms for computing the Nash equilibrium with complete information about the game, studies on Nash equilibrium in black-box games are less common. In this paper, we focus on learning the Nash equilibrium when the only available information about an agent's payoff comes in the form of empirical queries. We provide a no-regret learning algorithm that utilizes Gaussian processes to identify the equilibrium in such games. Our approach not only ensures a theoretical convergence rate but also demonstrates effectiveness across a variety collection of games through experimental validation.

arXiv:2405.06008

Wilsonian Renormalization of Neural Network Gaussian Processes

Separating relevant and irrelevant information is key to any modeling process or scientific inquiry. Theoretical physics offers a powerful tool for achieving this in the form of the renormalization group (RG). Here we demonstrate a practical approach to performing Wilsonian RG in the context of Gaussian Process (GP) Regression. We systematically integrate out the unlearnable modes of the GP kernel, thereby obtaining an RG flow of the GP in which the data sets the IR scale. In simple cases, this results in a universal flow of the ridge parameter, which becomes input-dependent in the richer scenario in which non-Gaussianities are included. In addition to being analytically tractable, this approach goes beyond structural analogies between RG and neural networks by providing a natural connection between RG flow and learnable vs. unlearnable modes. Studying such flows may improve our understanding of feature learning in deep neural networks, and enable us to identify potential universality classes in these models.

arXiv:2405.05424

Latent Variable Double Gaussian Process Model for Decoding Complex Neural Data

Non-parametric models, such as Gaussian Processes (GP), show promising results in the analysis of complex data. Their applications in neuroscience data have recently gained traction. In this research, we introduce a novel neural decoder model built upon GP models. The core idea is that two GPs generate neural data and their associated labels using a set of low- dimensional latent variables. Under this modeling assumption, the latent variables represent the underlying manifold or essential features present in the neural data. When GPs are trained, the latent variable can be inferred from neural data to decode the labels with a high accuracy. We demonstrate an application of this decoder model in a verbal memory experiment dataset and show that the decoder accuracy in predicting stimulus significantly surpasses the state-of-the-art decoder models. The preceding performance of this model highlights the importance of utilizing non-parametric models in the analysis of neuroscience data.

arXiv:2405.01365공식 코드 · Papers with Code

Dynamic Online Ensembles of Basis Expansions

Practical Bayesian learning often requires (1) online inference, (2) dynamic models, and (3) ensembling over multiple different models. Recent advances have shown how to use random feature approximations to achieve scalable, online ensembling of Gaussian processes with desirable theoretical properties and fruitful applications. One key to these methods' success is the inclusion of a random walk on the model parameters, which makes models dynamic. We show that these methods can be generalized easily to any basis expansion model and that using alternative basis expansions, such as Hilbert space Gaussian processes, often results in better performance. To simplify the process of choosing a specific basis expansion, our method's generality also allows the ensembling of several entirely different models, for example, a Gaussian process and polynomial regression. Finally, we propose a novel method to ensemble static and dynamic models together.

arXiv:2404.18650

Enhancing RSS-Based Visible Light Positioning by Optimal Calibrating the LED Tilt and Gain

This paper presents an optimal calibration scheme and a weighted least squares (LS) localization algorithm for received signal strength (RSS) based visible light positioning (VLP) systems, focusing on the often overlooked impact of light emitting diode (LED) tilt. By optimally calibrating LED tilt and gain, we significantly enhance VLP localization accuracy. Our algorithm outperforms both machine learning Gaussian processes (GPs) and traditional multilateration techniques. Against GPs, it achieves improvements of 58% and 74% in the 50th and 99th percentiles, respectively. When compared to multilateration, it reduces the 50th percentile error from 7.4 cm to 3.2 cm and the 99th percentile error from 25.7 cm to 11 cm. We introduce a low-complexity estimator for tilt and gain that meets the Cramer-Rao lower bound (CRLB) for the mean squared error (MSE), emphasizing its precision and efficiency. Further, we elaborate on optimal calibration measurement placement and refine the observation model to include residual calibration errors, thereby improving localization performance. The weighted LS algorithm's effectiveness is validated through simulations and real-world data, consistently outperforming GPs and multilateration, across various training set sizes and reducing outlier errors. Our findings underscore the critical role of LED tilt calibration in advancing VLP system accuracy and contribute to a more precise model for indoor positioning technologies.

arXiv:2404.19157

Scalable Bayesian Inference in the Era of Deep Learning: From Gaussian Processes to Deep Neural Networks

Large neural networks trained on large datasets have become the dominant paradigm in machine learning. These systems rely on maximum likelihood point estimates of their parameters, precluding them from expressing model uncertainty. This may result in overconfident predictions and it prevents the use of deep learning models for sequential decision making. This thesis develops scalable methods to equip neural networks with model uncertainty. In particular, we leverage the linearised Laplace approximation to equip pre-trained neural networks with the uncertainty estimates provided by their tangent linear models. This turns the problem of Bayesian inference in neural networks into one of Bayesian inference in conjugate Gaussian-linear models. Alas, the cost of this remains cubic in either the number of network parameters or in the number of observations times output dimensions. By assumption, neither are tractable. We address this intractability by using stochastic gradient descent (SGD) -- the workhorse algorithm of deep learning -- to perform posterior sampling in linear models and their convex duals: Gaussian processes. With this, we turn back to linearised neural networks, finding the linearised Laplace approximation to present a number of incompatibilities with modern deep learning practices -- namely, stochastic optimisation, early stopping and normalisation layers -- when used for hyperparameter learning. We resolve these and construct a sample-based EM algorithm for scalable hyperparameter learning with linearised neural networks. We apply the above methods to perform linearised neural network inference with ResNet-50 (25M parameters) trained on Imagenet (1.2M observations and 1000 output dimensions). Additionally, we apply our methods to estimate uncertainty for 3d tomographic reconstructions obtained with the deep image prior network.