Label-Free Distribution Shift Detection in Truck Mode Classification via Uncertainty-Gap Monitors

Bek-Myrza Nurmatov, Valery Manokhin, Jens Heger
Proceedings of the Fifteenth Symposium on Conformal and Probabilistic Prediction with Applications, PMLR 329:1087-1089, 2026.

Abstract

Machine learning models in vehicle telematics face distribution shifts, but existing sequential monitors often require real-time ground-truth labels. We introduce the Uncertainty-Gap Monitor (UGM), a label-free empirical monitor based on the width of Venn-Abers probability intervals. In truck accelerometer data, UGM remained stable on benign shifts and produced large e-values that triggered alarms on the aggregated out-of-distribution fleet shifts.

Cite this Paper

BibTeX
@InProceedings{pmlr-v329-nurmatov26a, title = {Label-Free Distribution Shift Detection in Truck Mode Classification via Uncertainty-Gap Monitors}, author = {Nurmatov, Bek-Myrza and Manokhin, Valery and Heger, Jens}, booktitle = {Proceedings of the Fifteenth Symposium on Conformal and Probabilistic Prediction with Applications}, pages = {1087--1089}, year = {2026}, editor = {Ahlberg, Ernst and Johansson, Ulf and Boström, Henrik and Carlevaro, Alberto and Hallberg Szabadváry, Johan and Carlsson, Lars}, volume = {329}, series = {Proceedings of Machine Learning Research}, month = {02--04 Sep}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v329/main/assets/nurmatov26a/nurmatov26a.pdf}, url = {https://proceedings.mlr.press/v329/nurmatov26a.html}, abstract = {Machine learning models in vehicle telematics face distribution shifts, but existing sequential monitors often require real-time ground-truth labels. We introduce the Uncertainty-Gap Monitor (UGM), a label-free empirical monitor based on the width of Venn-Abers probability intervals. In truck accelerometer data, UGM remained stable on benign shifts and produced large e-values that triggered alarms on the aggregated out-of-distribution fleet shifts.} }
Endnote
%0 Conference Paper %T Label-Free Distribution Shift Detection in Truck Mode Classification via Uncertainty-Gap Monitors %A Bek-Myrza Nurmatov %A Valery Manokhin %A Jens Heger %B Proceedings of the Fifteenth Symposium on Conformal and Probabilistic Prediction with Applications %C Proceedings of Machine Learning Research %D 2026 %E Ernst Ahlberg %E Ulf Johansson %E Henrik Boström %E Alberto Carlevaro %E Johan Hallberg Szabadváry %E Lars Carlsson %F pmlr-v329-nurmatov26a %I PMLR %P 1087--1089 %U https://proceedings.mlr.press/v329/nurmatov26a.html %V 329 %X Machine learning models in vehicle telematics face distribution shifts, but existing sequential monitors often require real-time ground-truth labels. We introduce the Uncertainty-Gap Monitor (UGM), a label-free empirical monitor based on the width of Venn-Abers probability intervals. In truck accelerometer data, UGM remained stable on benign shifts and produced large e-values that triggered alarms on the aggregated out-of-distribution fleet shifts.
APA
Nurmatov, B., Manokhin, V. & Heger, J.. (2026). Label-Free Distribution Shift Detection in Truck Mode Classification via Uncertainty-Gap Monitors. Proceedings of the Fifteenth Symposium on Conformal and Probabilistic Prediction with Applications, in Proceedings of Machine Learning Research 329:1087-1089 Available from https://proceedings.mlr.press/v329/nurmatov26a.html.

Related Material