Multi-model ensemble analysis integrates information from multiple climate models into a unified projection. However, existing integration approaches based on model averaging can dilute fine-scale spatial information and incur bias from rescaling low-resolution climate models. We propose a statistical approach, called NN-GPR, using Gaussian process regression (GPR) with an infinitely wide deep neural network based covariance function. NN-GPR requires no assumptions about the relationships between models, no interpolation to a common grid, no stationarity assumptions, and automatically downscales as part of its prediction algorithm. Model experiments show that NN-GPR can be highly skillful at surface temperature and precipitation forecasting by preserving geospatial signals at multiple scales and capturing inter-annual variability. Our projections particularly show improved accuracy and uncertainty quantification skill in regions of high variability, which allows us to cheaply assess tail behavior at a 0.44$^\circ$/50 km spatial resolution without a regional climate model (RCM). Evaluations on reanalysis data and SSP245 forced climate models show that NN-GPR produces similar, overall climatologies to the model ensemble while better capturing fine scale spatial patterns. Finally, we compare NN-GPR's regional predictions against two RCMs and show that NN-GPR can rival the performance of RCMs using only global model data as input.
Gaussian process regression (GPR) is a fundamental model used in machine learning (ML). Due to its accurate prediction with uncertainty and versatility in handling various data structures via kernels, GPR has been successfully used in various applications. However, in GPR, how the features of an input contribute to its prediction cannot be interpreted. Here, we propose GPR with local explanation, which reveals the feature contributions to the prediction of each sample while maintaining the predictive performance of GPR. In the proposed model, both the prediction and explanation for each sample are performed using an easy-to-interpret locally linear model. The weight vector of the locally linear model is assumed to be generated from multivariate Gaussian process priors. The hyperparameters of the proposed models are estimated by maximizing the marginal likelihood. For a new test sample, the proposed model can predict the values of its target variable and weight vector, as well as their uncertainties, in a closed form. Experimental results on various benchmark datasets verify that the proposed model can achieve predictive performance comparable to those of GPR and superior to that of existing interpretable models and can achieve higher interpretability than them, both quantitatively and qualitatively.
Designing spectral convolutional networks is a challenging problem in graph learning. ChebNet, one of the early attempts, approximates the spectral graph convolutions using Chebyshev polynomials. GCN simplifies ChebNet by utilizing only the first two Chebyshev polynomials while still outperforming it on real-world datasets. GPR-GNN and BernNet demonstrate that the Monomial and Bernstein bases also outperform the Chebyshev basis in terms of learning the spectral graph convolutions. Such conclusions are counter-intuitive in the field of approximation theory, where it is established that the Chebyshev polynomial achieves the optimum convergent rate for approximating a function. In this paper, we revisit the problem of approximating the spectral graph convolutions with Chebyshev polynomials. We show that ChebNet's inferior performance is primarily due to illegal coefficients learnt by ChebNet approximating analytic filter functions, which leads to over-fitting. We then propose ChebNetII, a new GNN model based on Chebyshev interpolation, which enhances the original Chebyshev polynomial approximation while reducing the Runge phenomenon. We conducted an extensive experimental study to demonstrate that ChebNetII can learn arbitrary graph convolutions and achieve superior performance in both full- and semi-supervised node classification tasks. Most notably, we scale ChebNetII to a billion graph ogbn-papers100M, showing that spectral-based GNNs have superior performance. Our code is available at https://github.com/ivam-he/ChebNetII.
We consider the problem of estimating a good maximizer of a black-box function given noisy examples. To solve such problems, we propose to fit a new type of function which we call a global optimization network (GON), defined as any composition of an invertible function and a unimodal function, whose unique global maximizer can be inferred in $\mathcal{O}(D)$ time. In this paper, we show how to construct invertible and unimodal functions by using linear inequality constraints on lattice models. We also extend to \emph{conditional} GONs that find a global maximizer conditioned on specified inputs of other dimensions. Experiments show the GON maximizers are statistically significantly better predictions than those produced by convex fits, GPR, or DNNs, and are more reasonable predictions for real-world problems.
With the rapid expansion of urban areas and the increasingly use of electricity, the need for locating buried cables is becoming urgent. In this paper, a noval method to locate underground cables based on Ground Penetrating Radar (GPR) and Gaussian-process regression is proposed. Firstly, the coordinate system of the detected area is conducted, and the input and output of locating buried cables are determined. The GPR is moved along the established parallel detection lines, and the hyperbolic signatures generated by buried cables are identified and fitted, thus the positions and depths of some points on the cable could be derived. On the basis of the established coordinate system and the derived points on the cable, the clustering method and cable fitting algorithm based on Gaussian-process regression are proposed to find the most likely locations of the underground cables. Furthermore, the confidence intervals of the cable's locations are also obtained. Both the position and depth noises are taken into account in our method, ensuring the robustness and feasibility in different environments and equipments. Experiments on real-world datasets are conducted, and the obtained results demonstrate the effectiveness of the proposed method.
In this thesis a probabilistic framework is developed and proposed for Dynamic Object Recognition in 3D Environments. A software package is developed using C++ and Python in ROS that performs the detection and tracking task. Furthermore, a novel Gaussian Process Regression (GPR) based method is developed to detect ground points in different urban scenarios of regular, sloped and rough. The ground surface behavior is assumed to only demonstrate local input-dependent smoothness. kernel's length-scales are obtained. Bayesian inference is implemented sing \textit{Maximum a Posteriori} criterion. The log-marginal likelihood function is assumed to be a multi-task objective function, to represent a whole-frame unbiased view of the ground at each frame because adjacent segments may not have similar ground structure in an uneven scene while having shared hyper-parameter values. Simulation results shows the effectiveness of the proposed method in uneven and rough scenes which outperforms similar Gaussian process based ground segmentation methods.
Ground Penetrating Radar (GPR) is widely used as a non-destructive approach to estimate buried utilities. When the GPR's detecting direction is perpendicular to a pipeline, a hyperbolic characteristic would be formed on the GPR B-scan image. However, in real-world applications, the direction of pipelines on the existing pipeline map could be inaccurate, and it is hard to ensure the moving direction of GPR to be actually perpendicular to underground pipelines. In this paper, a novel model is proposed to estimate the direction and radius of pipeline and revise the existing pipeline map from GPR B-scan images. The model consists of two parts: GPR B-scan image processing and Ellipse Iterative Inversion Algorithm (EIIA). Firstly, the GPR B-scan image is processed with downward-opening point set extracted. The obtained point set is then iteratively inverted to the elliptical cross section of the buried pipeline, which is caused by the angle between the GPR's detecting direction and the pipeline's direction. By minimizing the sum of the algebraic distances from the extracted point set to the inverted ellipse, the most likely pipeline's direction and radius are determined. Experiments on real-world datasets are conducted, and the results demonstrate the effectiveness of the method.
Outlier detection of semiconductor devices is important since manufacturing variation is inherently inevitable. In order to properly detect outliers, it is necessary to consider the discrepancy from underlying trend. Conventional methods are insufficient as they cannot track spatial changes of the trend}}. This study proposes an adaptive outlier detection using Gaussian process regression (GPR) with Student-t likelihood, which captures a gradual spatial change of characteristic variation. According to the credible interval of the GPR posterior distribution, the devices having excessively large deviations against the underlying trend are detected. The proposed methodology is validated by the experiments using a commercial SiC wafer and simulation.
Plant phenotyping (Guo et al. 2021; Pieruschka et al. 2019) focuses on studying the diverse traits of plants related to the plants' growth. To be more specific, by accurately measuring the plant's anatomical, ontogenetical, physiological and biochemical properties, it allows identifying the crucial factors of plants' growth in different environments. One commonly used approach is to predict the plant's traits using hyperspectral reflectance (Yendrek et al. 2017; Wang et al. 2021). However, the data distributions of the hyperspectral reflectance data in plant phenotyping might vary in different environments for different plants. That is, it would be computationally expansive to learn the machine learning models separately for one plant in different environments. To solve this problem, we focus on studying the knowledge transferability of modern machine learning models in plant phenotyping. More specifically, this work aims to answer the following questions. (1) How is the performance of conventional machine learning models, e.g., partial least squares regression (PLSR), Gaussian process regression (GPR) and multi-layer perceptron (MLP), affected by the number of annotated samples for plant phenotyping? (2) Whether could the neural network based transfer learning models improve the performance of plant phenotyping? (3) Could the neural network based transfer learning be improved by using infinite-width hidden layers for plant phenotyping?
Ground-penetrating radar (GPR) has been used as a non-destructive tool for tree root inspection. Estimating root-related parameters from GPR radargrams greatly facilitates root health monitoring and imaging. However, the task of estimating root-related parameters is challenging as the root reflection is a complex function of multiple root parameters and root orientations. Existing methods can only estimate a single root parameter at a time without considering the influence of other parameters and root orientations, resulting in limited estimation accuracy under different root conditions. In addition, soil heterogeneity introduces clutter in GPR radargrams, making the data processing and interpretation even harder. To address these issues, a novel neural network architecture, called mask-guided multi-polarimetric integration neural network (MMI-Net), is proposed to automatically and simultaneously estimate multiple root-related parameters in heterogeneous soil environments. The MMI-Net includes two sub-networks: a MaskNet that predicts a mask to highlight the root reflection area to eliminate interfering environmental clutter, and a ParaNet that uses the predicted mask as guidance to integrate, extract, and emphasize informative features in multi-polarimetric radargrams for accurate estimation of five key root-related parameters. The parameters include the root depth, diameter, relative permittivity, horizontal and vertical orientation angles. Experimental results demonstrate that the proposed MMI-Net achieves high estimation accuracy in these root-related parameters. This is the first work that takes the combined contributions of root parameters and spatial orientations into account and simultaneously estimates multiple root-related parameters. The data and code implemented in the paper can be found at https://haihan-sun.github.io/GPR.html.
General full-wave electromagnetic solvers, such as those utilizing the finite-difference time-domain (FDTD) method, are computationally demanding for simulating practical GPR problems. We explore the performance of a near-real-time, forward modeling approach for GPR that is based on a machine learning (ML) architecture. To ease the process, we have developed a framework that is capable of generating these ML-based forward solvers automatically. The framework uses an innovative training method that combines a predictive dimensionality reduction technique and a large data set of modeled GPR responses from our FDTD simulation software, gprMax. The forward solver is parameterized for a specific GPR application, but the framework can be extended in a straightforward manner to different electromagnetic problems.
Learning-based control methods utilize run-time data from the underlying process to improve the controller performance under model mismatch and unmodeled disturbances. This is beneficial for optimizing industrial processes, where the dynamics are difficult to model, and the repetitive nature of the process can be exploited. In this work, we develop an iterative approach for repetitive precision motion control problems where the objective is to follow a reference geometry with minimal tracking error. Our method utilizes a nominal model of the process and learns the mismatch using Gaussian Process Regression (GPR). The control input and the GPR data are updated after each iteration to improve the performance in a run-to-run fashion. We provide a preliminary convergence analysis, implementation details of the proposed controller for minimizing different error types, and a case study where we demonstrate improved tracking performance with simulation and experimental results.
Current seismic design codes primarily rely on the strength and displacement capacity of structural members and do not account for the influence of the ground motion duration or the hysteretic behavior characteristics. The energy-based approach serves as a supplemental index to response quantities and includes the effect of repeated loads in seismic performance. The design philosophy suggests that the seismic demands are met by the energy dissipation capacity of the structural members. Therefore, the energy dissipation behavior of the structural members should be well understood to achieve an effective energy-based design approach. This study focuses on the energy dissipation capacity of reinforced concrete (RC) shear walls that are widely used in high seismic regions as they provide significant stiffness and strength to resist lateral forces. A machine learning (Gaussian Process Regression (GPR))-based predictive model for energy dissipation capacity of shear walls is developed as a function of wall design parameters. Eighteen design parameters are shown to influence energy dissipation, whereas the most important ones are determined by applying sequential backward elimination and by using feature selection methods to reduce the complexity of the predictive model. The ability of the proposed model to make robust and accurate predictions is validated based on novel data with a prediction accuracy (the ratio of predicted/actual values) of around 1.00 and a coefficient of determination (R2) of 0.93. The outcomes of this study are believed to contribute to the energy-based approach by (i) defining the most influential wall properties on the seismic energy dissipation capacity of shear walls and (ii) providing predictive models that can enable comparisons of different wall design configurations to achieve higher energy dissipation capacity.
Building Information Modeling (BIM) is increasingly used in the construction industry, but existing studies often ignore embedded rebars. Ground Penetrating Radar (GPR) provides a potential solution to develop as-built BIM with surface elements and rebars. However, automatically translating rebars from GPR into BIM is challenging since GPR cannot provide any information about the scanned element. Thus, we propose an approach to link GPR data and BIM according to Faster R-CNN. A label is attached to each element scanned by GPR for capturing the labeled images, which are used with other images to build a 3D model. Meanwhile, Faster R-CNN is introduced to identify the labels, and the projection relationship between images and the model is used to localize the scanned elements in the 3D model. Two concrete buildings is selected to evaluate the proposed approach, and the results reveal that our method could accurately translate the rebars from GPR data into corresponding elements in BIM with correct distributions.
We characterize the power-law asymptotics of learning curves for Gaussian process regression (GPR) under the assumption that the eigenspectrum of the prior and the eigenexpansion coefficients of the target function follow a power law. Under similar assumptions, we leverage the equivalence between GPR and kernel ridge regression (KRR) to show the generalization error of KRR. Infinitely wide neural networks can be related to GPR with respect to the neural network GP kernel and the neural tangent kernel, which in several cases is known to have a power-law spectrum. Hence our methods can be applied to study the generalization error of infinitely wide neural networks. We present toy experiments demonstrating the theory.
Ubiquitous use of lithium-ion batteries across multiple industries presents an opportunity to explore cost saving initiatives as the price to performance ratio continually decreases in a competitive environment. Manufacturers using lithium-ion batteries ranging in applications from mobile phones to electric vehicles need to know how long batteries will last for a given service life. To understand this, expensive testing is required. This paper utilizes the data and methods implemented by Kristen A. Severson, et al, to explore the methodologies that the research team used and presents another method to compare predicted results vs. actual test data for battery capacity fade. The fundamental effort is to find out if machine learning techniques may be trained to use early life cycle data in order to accurately predict battery capacity over the battery life cycle. Results show comparison of methods between Gaussian Process Regression (GPR) and Elastic Net Regression (ENR) and highlight key data features used from the extensive dataset found in the work of Severson, et al.
Context: Modelling satellite galaxy abundance $N_s$ in Galaxy Clusters (GCs) is a key element in modelling the Halo Occupation Distribution (HOD), which itself is a powerful tool to connect observational studies with numerical simulations. Aims: To study the impact of cosmological parameters on satellite abundance both in cosmological simulations and in mock observations. Methods: We build an emulator (HODEmu, \url{https://github.com/aragagnin/HODEmu/}) of satellite abundance based on cosmological parameters $\Omega_m, \Omega_b, \sigma_8, h_0$ and redshift $z.$ We train our emulator using \magneticum hydrodynamic simulations that span 15 different cosmologies, each over $4$ redshift slices between $0<z<0.5,$ and for each setup we fit normalisation $A$, log-slope $\beta$ and Gaussian fractional-scatter $\sigma$ of the $N_s-M$ relation. The emulator is based on multi-variate output Gaussian Process Regression (GPR). Results: We find that $A$ and $\beta$ depend on cosmological parameters, even if weakly, especially on $\Omega_m,$ $\Omega_b.$ This dependency can explain some discrepancies found in literature between satellite HOD of different cosmological simulations (Magneticum, Illustris, BAHAMAS). We also show that satellite abundance cosmology dependency differs between full-physics (FP) simulations, dark-matter only (DMO), and non-radiative simulations. Conclusions: This work provides a preliminary calibration of the cosmological dependency of the satellite abundance of high mass halos, and we showed that modelling HOD with cosmological parameters is necessary to interpret satellite abundance, and we showed the importance of using FP simulations in modelling this dependency.
Normative modelling is an emerging method for understanding the underlying heterogeneity within brain disorders like Alzheimer Disease (AD) by quantifying how each patient deviates from the expected normative pattern that has been learned from a healthy control distribution. Since AD is a multifactorial disease with more than one biological pathways, multimodal magnetic resonance imaging (MRI) neuroimaging data can provide complementary information about the disease heterogeneity. However, existing deep learning based normative models on multimodal MRI data use unimodal autoencoders with a single encoder and decoder that may fail to capture the relationship between brain measurements extracted from different MRI modalities. In this work, we propose multi-modal variational autoencoder (mmVAE) based normative modelling framework that can capture the joint distribution between different modalities to identify abnormal brain structural patterns in AD. Our multi-modal framework takes as input Freesurfer processed brain region volumes from T1-weighted (cortical and subcortical) and T2-weighed (hippocampal) scans of cognitively normal participants to learn the morphological characteristics of the healthy brain. The estimated normative model is then applied on Alzheimer Disease (AD) patients to quantify the deviation in brain volumes and identify the abnormal brain structural patterns due to the effect of the different AD stages. Our experimental results show that modeling joint distribution between the multiple MRI modalities generates deviation maps that are more sensitive to disease staging within AD, have a better correlation with patient cognition and result in higher number of brain regions with statistically significant deviations compared to a unimodal baseline model with all modalities concatenated as a single input.
We address a three-tier numerical framework based on manifold learning for the forecasting of high-dimensional time series. At the first step, we embed the time series into a reduced low-dimensional space using a nonlinear manifold learning algorithm such as Locally Linear Embedding and Diffusion Maps. At the second step, we construct reduced-order regression models on the manifold, in particular Multivariate Autoregressive (MVAR) and Gaussian Process Regression (GPR) models, to forecast the embedded dynamics. At the final step, we lift the embedded time series back to the original high-dimensional space using Radial Basis Functions interpolation and Geometric Harmonics. For our illustrations, we test the forecasting performance of the proposed numerical scheme with four sets of time series: three synthetic stochastic ones resembling EEG signals produced from linear and nonlinear stochastic models with different model orders, and one real-world data set containing daily time series of 10 key foreign exchange rates (FOREX) spanning the time period 03/09/2001-29/10/2020. The forecasting performance of the proposed numerical scheme is assessed using the combinations of manifold learning, modelling and lifting approaches. We also provide a comparison with the Principal Component Analysis algorithm as well as with the naive random walk model and the MVAR and GPR models trained and implemented directly in the high-dimensional space.
AI-assisted molecular optimization is a very active research field as it is expected to provide the next-generation drugs and molecular materials. An important difficulty is that the properties to be optimized rely on costly evaluations. Machine learning methods are investigated with success to predict these properties, but show generalization issues on less known areas of the chemical space. We propose here a surrogate-based black box optimization method, to tackle jointly the optimization and machine learning problems. It consists in optimizing the expected improvement of the surrogate of a molecular property using an evolutionary algorithm. The surrogate is defined as a Gaussian Process Regression (GPR) model, learned on a relevant area of the search space with respect to the property to be optimized. We show that our approach can successfully optimize a costly property of interest much faster than a purely metaheuristic approach.