Transferring model structure in Bayesian transfer learning for Gaussian process regression

Papež, Milan; Quinn, Anthony G.

doi:10.48550/arxiv.2101.06884

Cited by 2 publications

(5 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…• In future work, the target can itself be a joint network modeller, independent of the source(s), as in [18], further enhancing the positive transfer.…”

Section: Discussionmentioning

confidence: 99%

“…Model ( 14), ( 15), ( 16), together with (18), defines the linear state-space mode with uniform additive noises on orthotopic supports, denote the LSU-UOS model. Its observation and state evolution models (1) are equivalently specified as…”

Section: Lsu-uos Filtering Task For the Isolated Targetmentioning

confidence: 99%

“…at an infinite filtering horizon and so cannot be implemented (the curse of dimensionality [35]). In [41,42], approximate Bayesian filtering with the LSU model ( 19) and (20), closed within the UOS class (18), is proposed. This involves a local approximation after each data update (2) and time update (3), as recalled below.…”

Section: Lsu-uos Filtering Task For the Isolated Targetmentioning

confidence: 99%

“…As already noted, these methods require a complete model of source-target interaction. In contrast, in [18], BTL is defined as the task of conditioning a target probability distribution on a transferred source distribution. A dualmodeller framework is adopted, where the target modeller conditions on a probabilistic data predictor provided by an independent local source modeller.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Fully probabilistic design for knowledge fusion between Bayesian filters under uniform disturbances

Pavelková¹,

Jirsa²,

Quinn³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

This paper considers the problem of Bayesian transfer learning-based knowledge fusion between linear state-space processes driven by uniform state and observation noise processes. The target task conditions on probabilistic state predictor(s) supplied by the source filtering task(s) to improve its own state estimate. A joint model of the target and source(s) is not required and is not elicited. The resulting decision-making problem for choosing the optimal conditional target filtering distribution under incomplete modelling is solved via fully probabilistic design (FPD), i.e. via appropriate minimization of Kullback-Leibler divergence (KLD). The resulting FPD-optimal target learner is robust, in the sense that it can reject poor-quality source knowledge. In addition, the fact that this Bayesian transfer learning (BTL) scheme does not depend on a model of interaction between the source and target tasks ensures robustness to the misspecification of such a model. The latter is a problem that affects conventional transfer learning methods. The properties of the proposed BTL scheme are demonstrated via extensive simulations, and in comparison with two contemporary alternatives.

show abstract

“…• In future work, the target can itself be a joint network modeller, independent of the source(s), as in [18], further enhancing the positive transfer.…”

Section: Discussionmentioning

confidence: 99%

Section: Lsu-uos Filtering Task For the Isolated Targetmentioning

confidence: 99%

Section: Lsu-uos Filtering Task For the Isolated Targetmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Fully probabilistic design for knowledge fusion between Bayesian filters under uniform disturbances

Pavelková¹,

Jirsa²,

Quinn³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Particularly, in order to account for the asymmetric scenario raised from for example multi-fidelity modeling [16] and transfer learning [17], the asymmetric modeling structure has been investigated in the MTGP paradigm [18]- [21]. The MTGPs have also been studied in other regimes, like transfer learning [22] and few-shot learning [23].…”

Section: Introductionmentioning

confidence: 99%

Learning Multi-Task Gaussian Process Over Heterogeneous Input Domains

Liu¹,

Wu²,

Ong³

et al. 2022

Preprint

View full text Add to dashboard Cite

Multi-task Gaussian process (MTGP) is a wellknown non-parametric Bayesian model for learning correlated tasks effectively by transferring knowledge across tasks. But current MTGP models are usually limited to the multi-task scenario defined in the same input domain, leaving no space for tackling the practical heterogeneous case, i.e., the features of input domains vary over tasks. To this end, this paper presents a novel heterogeneous stochastic variational linear model of coregionalization (HSVLMC) model for simultaneously learning the tasks with varied input domains. Particularly, we develop the stochastic variational framework with a Bayesian calibration method that (i) takes into account the effect of dimensionality reduction raised by domain mapping in order to achieve effective input alignment; and (ii) employs a residual modeling strategy to leverage the inductive bias brought by prior domain mappings for better model inference. Finally, the superiority of the proposed model against existing LMC models has been extensively verified on diverse heterogeneous multi-task cases.

show abstract

Transferring model structure in Bayesian transfer learning for Gaussian process regression

Cited by 2 publications

References 24 publications

Fully probabilistic design for knowledge fusion between Bayesian filters under uniform disturbances

Fully probabilistic design for knowledge fusion between Bayesian filters under uniform disturbances

Learning Multi-Task Gaussian Process Over Heterogeneous Input Domains

Contact Info

Product

Resources

About