Partially Exchangeable Networks and Architectures for Learning Summary Statistics in Approximate Bayesian Computation

Samuel Wiqvist, Pierre-Alexandre Mattei, Umberto Picchini, Jes Frellsen

Publikation: Konference artikel i Proceeding eller bog/rapport kapitelKonferencebidrag i proceedingsForskningpeer review

Abstrakt

We present a novel family of deep neural architectures, named partially exchangeable networks (PENs) that leverage probabilistic symmetries. By design, PENs are invariant to block-switch transformations, which characterize the partial exchangeability properties of conditionally Markovian processes. Moreover, we show that any block-switch invariant function has a PEN-like representation. The DeepSets architecture is a special case of PEN and we can therefore also target fully exchangeable data. We employ PENs to learn summary statistics in approximate Bayesian computation (ABC). When comparing PENs to previous deep learning methods for learning summary statistics, our results are highly competitive, both considering time series and static models. Indeed, PENs provide more reliable posterior samples even when using less training data.
OriginalsprogEngelsk
TitelProceedings of the 36th International Conference on Machine Learning, PMLR
Vol/bind97
Publikationsdato2019
Sider6798-6807
StatusUdgivet - 2019

Fingeraftryk

Dyk ned i forskningsemnerne om 'Partially Exchangeable Networks and Architectures for Learning Summary Statistics in Approximate Bayesian Computation'. Sammen danner de et unikt fingeraftryk.

Citationsformater