Scalable approximate Bayesian inference for particle tracking data

Ruoxi Sun, Liam Paninski
Proceedings of the 35th International Conference on Machine Learning, PMLR 80:4800-4809, 2018.

Abstract

Many important datasets in physics, chemistry, and biology consist of noisy sequences of images of multiple moving overlapping particles. In many cases, the observed particles are indistinguishable, leading to unavoidable uncertainty about nearby particles’ identities. Exact Bayesian inference is intractable in this setting, and previous approximate Bayesian methods scale poorly. Non-Bayesian approaches that output a single “best” estimate of the particle tracks (thus discarding important uncertainty information) are therefore dominant in practice. Here we propose a flexible and scalable amortized approach for Bayesian inference on this task. We introduce a novel neural network method to approximate the (intractable) filter-backward-sample-forward algorithm for Bayesian inference in this setting. By varying the simulated training data for the network, we can perform inference on a wide variety of data types. This approach is therefore highly flexible and improves on the state of the art in terms of accuracy; provides uncertainty estimates about the particle locations and identities; and has a test run-time that scales linearly as a function of the data length and number of particles, thus enabling Bayesian inference in arbitrarily large particle tracking datasets.

Cite this Paper

BibTeX
@InProceedings{pmlr-v80-sun18b, title = {Scalable approximate {B}ayesian inference for particle tracking data}, author = {Sun, Ruoxi and Paninski, Liam}, booktitle = {Proceedings of the 35th International Conference on Machine Learning}, pages = {4800--4809}, year = {2018}, editor = {Dy, Jennifer and Krause, Andreas}, volume = {80}, series = {Proceedings of Machine Learning Research}, month = {10--15 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v80/sun18b/sun18b.pdf}, url = {https://proceedings.mlr.press/v80/sun18b.html}, abstract = {Many important datasets in physics, chemistry, and biology consist of noisy sequences of images of multiple moving overlapping particles. In many cases, the observed particles are indistinguishable, leading to unavoidable uncertainty about nearby particles’ identities. Exact Bayesian inference is intractable in this setting, and previous approximate Bayesian methods scale poorly. Non-Bayesian approaches that output a single “best” estimate of the particle tracks (thus discarding important uncertainty information) are therefore dominant in practice. Here we propose a flexible and scalable amortized approach for Bayesian inference on this task. We introduce a novel neural network method to approximate the (intractable) filter-backward-sample-forward algorithm for Bayesian inference in this setting. By varying the simulated training data for the network, we can perform inference on a wide variety of data types. This approach is therefore highly flexible and improves on the state of the art in terms of accuracy; provides uncertainty estimates about the particle locations and identities; and has a test run-time that scales linearly as a function of the data length and number of particles, thus enabling Bayesian inference in arbitrarily large particle tracking datasets.} }
Endnote
%0 Conference Paper %T Scalable approximate Bayesian inference for particle tracking data %A Ruoxi Sun %A Liam Paninski %B Proceedings of the 35th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2018 %E Jennifer Dy %E Andreas Krause %F pmlr-v80-sun18b %I PMLR %P 4800--4809 %U https://proceedings.mlr.press/v80/sun18b.html %V 80 %X Many important datasets in physics, chemistry, and biology consist of noisy sequences of images of multiple moving overlapping particles. In many cases, the observed particles are indistinguishable, leading to unavoidable uncertainty about nearby particles’ identities. Exact Bayesian inference is intractable in this setting, and previous approximate Bayesian methods scale poorly. Non-Bayesian approaches that output a single “best” estimate of the particle tracks (thus discarding important uncertainty information) are therefore dominant in practice. Here we propose a flexible and scalable amortized approach for Bayesian inference on this task. We introduce a novel neural network method to approximate the (intractable) filter-backward-sample-forward algorithm for Bayesian inference in this setting. By varying the simulated training data for the network, we can perform inference on a wide variety of data types. This approach is therefore highly flexible and improves on the state of the art in terms of accuracy; provides uncertainty estimates about the particle locations and identities; and has a test run-time that scales linearly as a function of the data length and number of particles, thus enabling Bayesian inference in arbitrarily large particle tracking datasets.
APA
Sun, R. & Paninski, L.. (2018). Scalable approximate Bayesian inference for particle tracking data. Proceedings of the 35th International Conference on Machine Learning, in Proceedings of Machine Learning Research 80:4800-4809 Available from https://proceedings.mlr.press/v80/sun18b.html.

Related Material