• Corpus ID: 195345121

Multi-task Learning for Aggregated Data using Gaussian Processes

@article{Yousefi2019MultitaskLF,
  title={Multi-task Learning for Aggregated Data using Gaussian Processes},
  author={Fariba Yousefi and Michael Thomas Smith and Mauricio A {\'A}lvarez},
  journal={ArXiv},
  year={2019},
  volume={abs/1906.09412}
}
Aggregated data is commonplace in areas such as epidemiology and demography. For example, census data for a population is usually given as averages defined over time periods or spatial resolutions (cities, regions or countries). In this paper, we present a novel multi-task learning model based on Gaussian processes for joint learning of variables that have been aggregated at different input scales. Our model represents each task as the linear combination of the realizations of latent processes… 

Figures from this paper

Spatially Aggregated Gaussian Processes with Multivariate Areal Outputs

TLDR
A probabilistic model for inferring the multivariate function from multiple areal data sets with various granularities that can accurately refine coarse-grainedAreal data, and offer performance improvements by using the arealData sets from multiple domains.

Aggregated Multi-output Gaussian Processes with Knowledge Transfer Across Domains

TLDR
A multi-output Gaussian process (MoGP) model that infers functions for attributes using multiple aggregate datasets of respective granularities based on variational Bayes, which enables one to learn the model parameters using the aggregate datasets from multiple domains.

Probabilistic Models for Spatially Aggregated Data

TLDR
This thesis proposes probabilistic models and their inference algorithms that address the following two problems: 1) refining coarse-grained aggregated data; 2) inferring latent people flow from aggregated population data and demonstrates the effectiveness of the proposed models by extensive experiments on real-world data sets.

Multi-resolution Spatial Regression for Aggregated Data with an Application to Crop Yield Prediction

We develop a new methodology for spatial regression of aggregated outputs on multi-resolution covariates. Such problems often occur with spatial data, for example in crop yield prediction, where the

Aggregated Gaussian Processes with Multiresolution Earth Observation Covariates

TLDR
A Gaussian process regression model is proposed to handle multiresolution covariates by employing an additive kernel that can efficiently aggregate features across resolutions, leading to better performance and interpretability.

Multi-resolution Multi-task Gaussian Processes

TLDR
This work develops shallow Gaussian Process (GP) mixtures that approximate the difficult to estimate joint likelihood with a composite one and deep GP constructions that naturally handle biases in the mean and generalize and outperform state of the art GP compositions and offer information-theoretic corrections and efficient variational approximations.

Non-Linearities in Gaussian Processes with Integral Observations

TLDR
A solution based on Markov chain Monte Carlo with numerical integration for aggregation is presented and demonstrated in binned Poisson regression and in non-invasive detection of fouling using ultrasound waves.

Deconditional Downscaling with Gaussian Processes

TLDR
This work introduces conditional mean process (CMP), a new class of Gaussian Processes describing conditional means, and demonstrates its proficiency in a synthetic and a real-world atmospheric field downscaling problem, showing substantial improvements over existing methods.

Distributed Bayesian Varying Coefficient Modeling Using a Gaussian Process Prior

TLDR
The empirical results show that the combination schemes that satisfy the theoretical assumptions, including the one in the AMC algorithm, have better nominal coverage, shorter credible intervals, smaller mean square errors, and higher effective sample size than their main competitors across diverse simulations and in a real data analysis.

Aggregate Learning for Mixed Frequency Data

TLDR
A Mixed-Frequency Aggregate Learning (MF-AGL) model is proposed that predicts economic indicators for the smaller areas in real-time and finds that the proposed model predicts the regional heterogeneity of the labor market condition and the rapidly changing economic status.

References

SHOWING 1-10 OF 43 REFERENCES

Spatially Aggregated Gaussian Processes with Multivariate Areal Outputs

TLDR
A probabilistic model for inferring the multivariate function from multiple areal data sets with various granularities that can accurately refine coarse-grainedAreal data, and offer performance improvements by using the arealData sets from multiple domains.

Variational Learning on Aggregate Outputs with Gaussian Processes

TLDR
This work develops a framework which extends to several types of likelihoods, including the Poisson model for aggregated count data, and applies it to the fine-scale spatial modelling of malaria incidence, with over 1 million observations.

Heterogeneous Multi-output Gaussian Process Prediction

TLDR
A novel extension of multi-output Gaussian processes for handling heterogeneous outputs that uses a covariance function with a linear model of coregionalisation form to obtain tractable variational bounds amenable to stochastic variational inference.

Multi-resolution Multi-task Gaussian Processes

TLDR
This work develops shallow Gaussian Process (GP) mixtures that approximate the difficult to estimate joint likelihood with a composite one and deep GP constructions that naturally handle biases in the mean and generalize and outperform state of the art GP compositions and offer information-theoretic corrections and efficient variational approximations.

Generalized Linear Models for Aggregated Data

TLDR
This paper addresses the scenario where features are provided at the individual level, but the target variables are only available as histogram aggregates or order statistics, and proposes a simple algorithm to estimate the model parameters and individual level inferences via alternating imputation and standard generalized linear model fitting.

Sparse Gaussian Processes using Pseudo-inputs

TLDR
It is shown that this new Gaussian process (GP) regression model can match full GP performance with small M, i.e. very sparse solutions, and it significantly outperforms other approaches in this regime.

Gaussian Processes for Big Data

TLDR
Stochastic variational inference for Gaussian process models is introduced and it is shown how GPs can be variationally decomposed to depend on a set of globally relevant inducing variables which factorize the model in the necessary manner to perform Variational inference.

Multi-task Gaussian Process Prediction

TLDR
A model that learns a shared covariance function on input-dependent features and a "free-form" covariance matrix over tasks allows for good flexibility when modelling inter-task dependencies while avoiding the need for large amounts of data for training.

Sparse Convolved Gaussian Processes for Multi-output Regression

TLDR
A sparse approximation approach for dependent output Gaussian processes (GP) using the convolution process formalism to establish dependencies between output variables, where each latent function is represented as a GP.

Gaussian Processes Multiple Instance Learning

TLDR
To circumvent the intractability of exact GP inference and learning incurred by the non-continuous max function, the soft-max approximation is suggested; second, the use of witness indicator variables optimized with a deterministic annealing schedule.