Active Learning for Transformer-Based Fault Diagnosis in 5G and Beyond Mobile Networks
Bibliographic record
Abstract
As 5G and beyond mobile networks evolve, their increasing complexity necessitates advanced, automated, and datadriven fault diagnosis methods. While traditional data-driven methods falter with modern network complexities, Transformer models have proven highly effective for fault diagnosis through their efficient processing of sequential and time-series data. However, these Transformer-based methods demand substantial labeled data, which is costly to obtain. To address the lack of labeled data, we propose a novel active learning (AL) approach designed for Transformer-based fault diagnosis, tailored to the time-series nature of network data. AL reduces the need for extensive labeled datasets by iteratively selecting the most informative samples for labeling. Our AL method exploits the interpretability of Transformers, using their attention weights to create dependency graphs that represent processing patterns of data points. By formulating a one-class novelty detection problem on these graphs, we identify whether an unlabeled sample is processed differently from labeled ones in the previous training cycle and designate novel samples for expert annotation. Extensive experiments on real-world datasets show that our AL method achieves higher F1-scores than state-of-the-art AL algorithms with 50% fewer labeled samples and surpasses existing methods by up to 150% in identifying samples related to unseen fault types.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.005 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".