Hierarchical Multiple-Instance Learning for Vehicle-Level Battery Anomaly Detection from Charging Data Streams
- 1Department of Electrical and Electronic Engineering Technology, University of Johannesburg, Johannesburg 2092, South Africa
Abstract
Vehicle-level battery labels paired with many charging records create a weakly supervised rare-event problem. We evaluated a two-level multiple-instance framework on 191 vehicles (184 normal and seven source-coded abnormal) from the public EVBattery archive, using 38,200 charging snippets. Nine normal-reference anomaly detectors were combined with fixed or learned vehicle-level pooling under six repetitions of five-fold vehicle-disjoint internal cross-validation. All preprocessing and detector fitting used training vehicles only; score mapping used aggregation-training vehicles; decision thresholds used a separate 13-14-vehicle validation partition containing only one or two abnormal vehicles. PCA reconstruction with 95th-percentile pooling produced the highest numerical repeat-mean average precision (0.342) among 46 evaluated pipelines. A leave-one-repeat-out selection-sensitivity diagnostic selected this pipeline in five of six analyses and yielded mean held-repeat average precision 0.318 ± 0.078; because all repeats reuse the same 191 vehicles, this is not an independent test estimate. After averaging six out-of-fold probabilities per vehicle, average precision was 0.371 and AUROC 0.796. Threshold-based metrics are exploratory: the 30 validation-derived thresholds ranged from 0.267 to 0.999, and 17 equalled 0.500. No pairwise comparison survived Holm correction, uncertainty was wide, and calibration was poor. The results identify PCA plus upper-tail pooling only as the highest numerical internal benchmark in this cohort, not as an unbiased estimate of a selected final model, confirmed physical diagnosis, or deployment-ready probability estimation.
Related
- QSARmil: A Package for Molecular Multi-Instance Machine Learning
- Thermodynamics-Informed Learning Captures Hierarchical Hydrogen-Bond Networks in Deep Eutectic Solvents
- Pre-Refinement Prediction of Dopant Site Occupancy in M-Type Hexaferrites via Hierarchical and Uncertainty-Aware Machine-Learning.
- HierDTA: Hierarchical Structure-Aware Representation Learning for Drug–Target Affinity Prediction
- Mapping Hierarchical Motifs in disorder-order transitions: learning structural motifs in supercooled liquids and crystals
BibTeX
@article{makama2026hierarchical,
title = {{Hierarchical Multiple-Instance Learning for Vehicle-Level Battery Anomaly Detection from Charging Data Streams}},
author = {Mavis Fortunate Makama and Bonginkosi A. Thango},
journal = {Batteries},
year = {2026},
month = oct,
doi = {10.3390/batteries12100431},
url = {https://www.mdpi.com/2313-0105/12/10/431},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Hierarchical Multiple-Instance Learning for Vehicle-Level Battery Anomaly Detection from Charging Data Streams
AU - Mavis Fortunate Makama
AU - Bonginkosi A. Thango
PY - 2026
DA - 2026/10/15
JO - Batteries
DO - 10.3390/batteries12100431
UR - https://www.mdpi.com/2313-0105/12/10/431
AB - Vehicle-level battery labels paired with many charging records create a weakly supervised rare-event problem. We evaluated a two-level multiple-instance framework on 191 vehicles (184 normal and seven source-coded abnormal) from the public EVBattery archive, using 38,200 charging snippets. Nine normal-reference anomaly detectors were combined with fixed or learned vehicle-level pooling under six repetitions of five-fold vehicle-disjoint internal cross-validation. All preprocessing and detector fitting used training vehicles only; score mapping used aggregation-training vehicles; decision thresholds used a separate 13-14-vehicle validation partition containing only one or two abnormal vehicles. PCA reconstruction with 95th-percentile pooling produced the highest numerical repeat-mean average precision (0.342) among 46 evaluated pipelines. A leave-one-repeat-out selection-sensitivity diagnostic selected this pipeline in five of six analyses and yielded mean held-repeat average precision 0.318 ± 0.078; because all repeats reuse the same 191 vehicles, this is not an independent test estimate. After averaging six out-of-fold probabilities per vehicle, average precision was 0.371 and AUROC 0.796. Threshold-based metrics are exploratory: the 30 validation-derived thresholds ranged from 0.267 to 0.999, and 17 equalled 0.500. No pairwise comparison survived Holm correction, uncertainty was wide, and calibration was poor. The results identify PCA plus upper-tail pooling only as the highest numerical internal benchmark in this cohort, not as an unbiased estimate of a selected final model, confirmed physical diagnosis, or deployment-ready probability estimation.
SN - 2313-0105
ER -