What to Check Next? Bayesian Networks and Rollout-Based Lookahead for Quicker Inspections

Sebastiaan J. J. Jans, Thijs van Ommen, Ad Feelders, Cor Veenman
Proceedings of the 13th International Conference on Probabilistic Graphical Models, PMLR 343:392-407, 2026.

Abstract

Regulatory inspections, such as labour inspections, take time and money. Time and money are always limited. Inspectors often work with checklists. It could save a lot of time if they could focus first on checklist items that are likely to be non-compliant with the law. We implement a method that can guide an inspector in the field by suggesting which item on the checklist they should inspect next, based on their answers to previous checklist items. To do this, we train a Bayesian Network model on filled checklist data from the Norwegian Labour Inspection Authority. We test a myopic policy that always picks the item that is most likely to be non-compliant against a policy that looks one step ahead. The lookahead policy computes for each candidate item the expected value ($\gamma$-discounted returns) of inspecting it under a fixed policy that no longer updates probabilities, also known as a rollout. The expected value under the fixed policy is cheaply computed in closed form. We find that this lookahead policy increases performance greatly over the myopic policy, halving the catching times of non-compliances from roughly $5$ to $2.7$ checks (out of a maximum of $14$). The Bayesian network is however slightly outperformed across the board by a Binary-Relevance Naive Bayes model intended as a baseline.

Cite this Paper

BibTeX
@InProceedings{pmlr-v343-jans26a, title = {What to Check Next? {Bayesian} Networks and Rollout-Based Lookahead for Quicker Inspections}, author = {Jans, Sebastiaan J. J. and {van Ommen}, Thijs and Feelders, Ad and Veenman, Cor}, booktitle = {Proceedings of the 13th International Conference on Probabilistic Graphical Models}, pages = {392--407}, year = {2026}, editor = {Camps-Valls, Gustau and Leonelli, Manuele and Varando, Gherardo}, volume = {343}, series = {Proceedings of Machine Learning Research}, month = {09--11 Sep}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v343/main/assets/jans26a/jans26a.pdf}, url = {https://proceedings.mlr.press/v343/jans26a.html}, abstract = {Regulatory inspections, such as labour inspections, take time and money. Time and money are always limited. Inspectors often work with checklists. It could save a lot of time if they could focus first on checklist items that are likely to be non-compliant with the law. We implement a method that can guide an inspector in the field by suggesting which item on the checklist they should inspect next, based on their answers to previous checklist items. To do this, we train a Bayesian Network model on filled checklist data from the Norwegian Labour Inspection Authority. We test a myopic policy that always picks the item that is most likely to be non-compliant against a policy that looks one step ahead. The lookahead policy computes for each candidate item the expected value ($\gamma$-discounted returns) of inspecting it under a fixed policy that no longer updates probabilities, also known as a rollout. The expected value under the fixed policy is cheaply computed in closed form. We find that this lookahead policy increases performance greatly over the myopic policy, halving the catching times of non-compliances from roughly $5$ to $2.7$ checks (out of a maximum of $14$). The Bayesian network is however slightly outperformed across the board by a Binary-Relevance Naive Bayes model intended as a baseline.} }
Endnote
%0 Conference Paper %T What to Check Next? Bayesian Networks and Rollout-Based Lookahead for Quicker Inspections %A Sebastiaan J. J. Jans %A Thijs van Ommen %A Ad Feelders %A Cor Veenman %B Proceedings of the 13th International Conference on Probabilistic Graphical Models %C Proceedings of Machine Learning Research %D 2026 %E Gustau Camps-Valls %E Manuele Leonelli %E Gherardo Varando %F pmlr-v343-jans26a %I PMLR %P 392--407 %U https://proceedings.mlr.press/v343/jans26a.html %V 343 %X Regulatory inspections, such as labour inspections, take time and money. Time and money are always limited. Inspectors often work with checklists. It could save a lot of time if they could focus first on checklist items that are likely to be non-compliant with the law. We implement a method that can guide an inspector in the field by suggesting which item on the checklist they should inspect next, based on their answers to previous checklist items. To do this, we train a Bayesian Network model on filled checklist data from the Norwegian Labour Inspection Authority. We test a myopic policy that always picks the item that is most likely to be non-compliant against a policy that looks one step ahead. The lookahead policy computes for each candidate item the expected value ($\gamma$-discounted returns) of inspecting it under a fixed policy that no longer updates probabilities, also known as a rollout. The expected value under the fixed policy is cheaply computed in closed form. We find that this lookahead policy increases performance greatly over the myopic policy, halving the catching times of non-compliances from roughly $5$ to $2.7$ checks (out of a maximum of $14$). The Bayesian network is however slightly outperformed across the board by a Binary-Relevance Naive Bayes model intended as a baseline.
APA
Jans, S.J.J., van Ommen, T., Feelders, A. & Veenman, C.. (2026). What to Check Next? Bayesian Networks and Rollout-Based Lookahead for Quicker Inspections. Proceedings of the 13th International Conference on Probabilistic Graphical Models, in Proceedings of Machine Learning Research 343:392-407 Available from https://proceedings.mlr.press/v343/jans26a.html.

Related Material