2023-04-11-bickfordsmith23a.md

title	software	abstract	section	layout	series	publisher	issn	id	month	tex_title	firstpage	lastpage	page	order	cycles	bibtex_author	author	date	address	container-title	volume	genre	issued	pdf	extras
Prediction-Oriented Bayesian Active Learning	https://github.com/fbickfordsmith/epig	Information-theoretic approaches to active learning have traditionally focused on maximising the information gathered about the model parameters, most commonly by optimising the BALD score. We highlight that this can be suboptimal from the perspective of predictive performance. For example, BALD lacks a notion of an input distribution and so is prone to prioritise data of limited relevance. To address this we propose the expected predictive information gain (EPIG), an acquisition function that measures information gain in the space of predictions rather than parameters. We find that using EPIG leads to stronger predictive performance compared with BALD across a range of datasets and models, and thus provides an appealing drop-in replacement.	Regular Papers	inproceedings	Proceedings of Machine Learning Research	PMLR	2640-3498	bickfordsmith23a	0	Prediction-Oriented Bayesian Active Learning	7331	7348	7331-7348	7331	false	Bickford Smith, Freddie and Kirsch, Andreas and Farquhar, Sebastian and Gal, Yarin and Foster, Adam and Rainforth, Tom	given family Freddie Bickford Smith given family Andreas Kirsch given family Sebastian Farquhar given family Yarin Gal given family Adam Foster given family Tom Rainforth	2023-04-11		Proceedings of The 26th International Conference on Artificial Intelligence and Statistics	206	inproceedings	date-parts 2023 4 11	https://proceedings.mlr.press/v206/bickfordsmith23a/bickfordsmith23a.pdf