Scalable Nonparametric Bayesian Inference on Point Processes with Gaussian Processes

Yves-Laurent Kom Samo, Stephen Roberts
Proceedings of the 32nd International Conference on Machine Learning, PMLR 37:2227-2236, 2015.

Abstract

In this paper we propose an efficient, scalable non-parametric Gaussian process model for inference on Poisson point processes. Our model does not resort to gridding the domain or to introducing latent thinning points. Unlike competing models that scale as O(n^3) over n data points, our model has a complexity O(nk^2) where k << n. We propose a MCMC sampler and show that the model obtained is faster, more accurate and generates less correlated samples than competing approaches on both synthetic and real-life data. Finally, we show that our model easily handles data sizes not considered thus far by alternate approaches.

Cite this Paper


BibTeX
@InProceedings{pmlr-v37-samo15, title = {Scalable Nonparametric Bayesian Inference on Point Processes with Gaussian Processes}, author = {Samo, Yves-Laurent Kom and Roberts, Stephen}, booktitle = {Proceedings of the 32nd International Conference on Machine Learning}, pages = {2227--2236}, year = {2015}, editor = {Bach, Francis and Blei, David}, volume = {37}, series = {Proceedings of Machine Learning Research}, address = {Lille, France}, month = {07--09 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v37/samo15.pdf}, url = {https://proceedings.mlr.press/v37/samo15.html}, abstract = {In this paper we propose an efficient, scalable non-parametric Gaussian process model for inference on Poisson point processes. Our model does not resort to gridding the domain or to introducing latent thinning points. Unlike competing models that scale as O(n^3) over n data points, our model has a complexity O(nk^2) where k << n. We propose a MCMC sampler and show that the model obtained is faster, more accurate and generates less correlated samples than competing approaches on both synthetic and real-life data. Finally, we show that our model easily handles data sizes not considered thus far by alternate approaches.} }
Endnote
%0 Conference Paper %T Scalable Nonparametric Bayesian Inference on Point Processes with Gaussian Processes %A Yves-Laurent Kom Samo %A Stephen Roberts %B Proceedings of the 32nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2015 %E Francis Bach %E David Blei %F pmlr-v37-samo15 %I PMLR %P 2227--2236 %U https://proceedings.mlr.press/v37/samo15.html %V 37 %X In this paper we propose an efficient, scalable non-parametric Gaussian process model for inference on Poisson point processes. Our model does not resort to gridding the domain or to introducing latent thinning points. Unlike competing models that scale as O(n^3) over n data points, our model has a complexity O(nk^2) where k << n. We propose a MCMC sampler and show that the model obtained is faster, more accurate and generates less correlated samples than competing approaches on both synthetic and real-life data. Finally, we show that our model easily handles data sizes not considered thus far by alternate approaches.
RIS
TY - CPAPER TI - Scalable Nonparametric Bayesian Inference on Point Processes with Gaussian Processes AU - Yves-Laurent Kom Samo AU - Stephen Roberts BT - Proceedings of the 32nd International Conference on Machine Learning DA - 2015/06/01 ED - Francis Bach ED - David Blei ID - pmlr-v37-samo15 PB - PMLR DP - Proceedings of Machine Learning Research VL - 37 SP - 2227 EP - 2236 L1 - http://proceedings.mlr.press/v37/samo15.pdf UR - https://proceedings.mlr.press/v37/samo15.html AB - In this paper we propose an efficient, scalable non-parametric Gaussian process model for inference on Poisson point processes. Our model does not resort to gridding the domain or to introducing latent thinning points. Unlike competing models that scale as O(n^3) over n data points, our model has a complexity O(nk^2) where k << n. We propose a MCMC sampler and show that the model obtained is faster, more accurate and generates less correlated samples than competing approaches on both synthetic and real-life data. Finally, we show that our model easily handles data sizes not considered thus far by alternate approaches. ER -
APA
Samo, Y.K. & Roberts, S.. (2015). Scalable Nonparametric Bayesian Inference on Point Processes with Gaussian Processes. Proceedings of the 32nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 37:2227-2236 Available from https://proceedings.mlr.press/v37/samo15.html.

Related Material