Learning to Acquire Information

Yewen Pu, Leslie Pack Kaelbling, Armando Solar-Lezama
Proceedings of the 33rd Conference on Uncertainty in Artificial Intelligence, PMLR R15:511-520, 2017.

Abstract

We consider the problem of diagnosis where a set of simple observations are used to in- fer a potentially complex hidden hypothesis. Finding the optimal subset of observations is intractable in general, thus we focus on the problem of active diagnosis, where the agent selects the next most-informative observation based on the results of previous observations. We show that under the assumption of uniform observation entropy, one can build an impli- cation model which directly predicts the out- come of the potential next observation condi- tioned on the results of past observations, and selects the observation with the maximum en- tropy. This approach enjoys reduced computa- tion complexity by bypassing the complicated hypothesis space, and can be trained on obser- vation data alone, learning how to query with- out knowledge of the hidden hypothesis.

Cite this Paper

BibTeX
@InProceedings{pmlr-vR15-pu17a, title = {Learning to Acquire Information}, author = {Pu, Yewen and Kaelbling, Leslie Pack and Solar-Lezama, Armando}, booktitle = {Proceedings of the 33rd Conference on Uncertainty in Artificial Intelligence}, pages = {511--520}, year = {2017}, editor = {Elidan, Gal and Kersting, Kristian}, volume = {R15}, series = {Proceedings of Machine Learning Research}, month = {11--15 Aug}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/r15/main/assets/pu17a/pu17a.pdf}, url = {https://proceedings.mlr.press/r15/pu17a.html}, abstract = {We consider the problem of diagnosis where a set of simple observations are used to in- fer a potentially complex hidden hypothesis. Finding the optimal subset of observations is intractable in general, thus we focus on the problem of active diagnosis, where the agent selects the next most-informative observation based on the results of previous observations. We show that under the assumption of uniform observation entropy, one can build an impli- cation model which directly predicts the out- come of the potential next observation condi- tioned on the results of past observations, and selects the observation with the maximum en- tropy. This approach enjoys reduced computa- tion complexity by bypassing the complicated hypothesis space, and can be trained on obser- vation data alone, learning how to query with- out knowledge of the hidden hypothesis.}, note = {Reissued by PMLR on 04 October 2026.} }
Endnote
%0 Conference Paper %T Learning to Acquire Information %A Yewen Pu %A Leslie Pack Kaelbling %A Armando Solar-Lezama %B Proceedings of the 33rd Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2017 %E Gal Elidan %E Kristian Kersting %F pmlr-vR15-pu17a %I PMLR %P 511--520 %U https://proceedings.mlr.press/r15/pu17a.html %V R15 %X We consider the problem of diagnosis where a set of simple observations are used to in- fer a potentially complex hidden hypothesis. Finding the optimal subset of observations is intractable in general, thus we focus on the problem of active diagnosis, where the agent selects the next most-informative observation based on the results of previous observations. We show that under the assumption of uniform observation entropy, one can build an impli- cation model which directly predicts the out- come of the potential next observation condi- tioned on the results of past observations, and selects the observation with the maximum en- tropy. This approach enjoys reduced computa- tion complexity by bypassing the complicated hypothesis space, and can be trained on obser- vation data alone, learning how to query with- out knowledge of the hidden hypothesis. %Z Reissued by PMLR on 04 October 2026.
APA
Pu, Y., Kaelbling, L.P. & Solar-Lezama, A.. (2017). Learning to Acquire Information. Proceedings of the 33rd Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research R15:511-520 Available from https://proceedings.mlr.press/r15/pu17a.html. Reissued by PMLR on 04 October 2026.

Related Material