DeepICMGP surrogate models multiple outputs efficiently.
problem Challenges in modeling dependencies between multiple outputs using traditional multi-output GPs.
method Introduces hierarchical coregionalization structures across layers in DGPs.
result Demonstrates competitive performance and active learning strategies.
E-LMC improves spatial field prediction accuracy by linearizing complex fields.
problem Predicting complex spatial fields with high accuracy and efficiency.
method Introducing an invertible neural network to linearize nonlinear spatial fields, enabling the use of LMC for nonlinear problems.
result Maximum improvement of about 40% over original LMC, outperforming other models.
Develops a scalable multi-task Gaussian process with neural embedding for improved performance.
problem High model complexity and limited model capability in multi-task Gaussian processes.
method Neural embedding of coregionalization, advanced variational inference, sparse approximation.
result Higher prediction quality and better generalization of the NSVLMC model.
MO-GP models fill gaps in biophysical data with across-domain info transfer.
problem Gap filling of biophysical parameters LAI and fAPAR over rice areas.
method Multi-output Gaussian Processes (MO-GP) based on Linear Model of Coregionalization (LMC).
result MO-GP models successfully predict biophysical variables even in high missing data regimes.
Multitask Gaussian process (MTGP) is powerful for joint learning of multiple tasks with complicated correlation patterns. However, due to the assembling of additive independent latent functions, all current MTGPs including the salient linear model of coregionalization (LMC) and convolution frameworks cannot effectively…
The non-storability of electricity makes it unique among commodity assets, and it is an important driver of its price behaviour in secondary financial markets. The instantaneous and continuous matching of power supply with demand is a key factor explaining its volatility. During periods of high demand, costlier generat…
IPGP framework improves psychological assessment by integrating shared and unique traits.
problem Tackles the debate on shared vs unique personality traits across individuals.
method Uses Gaussian process coregionalization model for non-Gaussian ordinal data, with stochastic variational inference for scalability.
result Improves prediction and estimation of individualized factor structures compared to existing methods.
Paper introduces a new model to handle multi-task learning across different input domains.
problem Learning correlated tasks across varying input domains.
method Develops a novel heterogeneous stochastic variational linear model of coregionalization (HSVLMC) for multi-task learning.
result The proposed model outperforms existing models in diverse multi-task scenarios.
Establishes connection between MTDNN and multitask GP, revealing weight correlation as key to task sharing.
problem Limited theoretical understanding of information sharing in MTDNN.
method Derives multitask GP kernels for MTDNN and MTBNN, showing shared hyper-parameters and last layer weights.
result Information sharing in MTDNN is due to weight correlation, not intermediate layer weights.
We generalize the log Gaussian Cox process (LGCP) framework to model multiple correlated point data jointly. The observations are treated as realizations of multiple LGCPs, whose log intensities are given by linear combinations of latent functions drawn from Gaussian process priors. The combination coefficients are als…
CMDE uses deep learning to estimate causal effects from complex data.
problem Handling complex data structures like images for causal effect estimation.
method Causal Multi-task Deep Ensemble (CMDE) framework.
result CMDE outperforms state-of-the-art methods across various datasets and tasks.
The paper tackles counterfactual inference with multioutput deep kernels in high-dimensional settings.
problem Performing counterfactual inference with observational data in high-dimensional settings with multiple actions and outcomes.
method The paper presents a general class of counterfactual multi-task deep kernels models based on Structural Causal Models (SCM) and Gaussian Processes.
result The models estimate causal effects and learn policies efficiently, scaling well with high dimensions.
Gaussian processes (GPs), or distributions over arbitrary functions in a continuous domain, can be generalized to the multi-output case: a linear model of coregionalization (LMC) is one approach. LMCs estimate and exploit correlations across the multiple outputs. While model estimation can be performed efficiently for …
A scalable MOGP model with stochastic variational inference for many outputs.
problem Efficiently modeling data from multiple sources with many outputs.
method Stochastic variational inference for Latent Variable MOGP (LV-MOGP).
result Computational complexity per iteration is independent of the number of outputs.
Multi-output Gaussian processes have received increasing attention during the last few years as a natural mechanism to extend the powerful flexibility of Gaussian processes to the setup of multiple output variables. The key point here is the ability to design kernel functions that allow exploiting the correlations betw…
Detecting anomalies in multivariate functional data using Bayesian nonparametric methods.
problem Detecting anomalies in functional data.
method Bayesian nonparametric approach with infinite mixture of multi-output Gaussian processes.
result Anomalous observations assigned to small mixture components.
Fleet control method improves sample efficiency in IoT environments.
problem Improving sample efficiency in reinforcement learning for fleet control.
method Coregionalized Gaussian Process Policy Iteration for knowledge transfer between fleet members.
result Significantly outperforms baseline approaches in terms of median and variance of results.
Efficiently predicts high-fidelity PDE solutions using multi-fidelity Gaussian processes.
problem Expensive high-fidelity solutions for PDEs on discretized domains.
method Multi-Fidelity High-Order Gaussian Process (MFHoGP) that integrates multi-fidelity examples and scales to large numbers of outputs.
result Significantly reduces the cost of high-fidelity PDE solutions through efficient Gaussian process modeling.
Extends GP regression to complex Helmholtz problems, improving wavefield inference in brain elastography.
problem Infer complex Helmholtz wavefields from sparse, noisy data.
method Operator-informed Gaussian processes, realifying complex operator into real blocks, using PDE residuals and boundary traces.
result Competitive with finite-difference and neural-network methods, reconstructs brain shear curl field with high correlation.