Log-Based Anomaly Detection Without Ground-Truth: Evaluating Weakly Supervised, Semi-Supervised, and Unsupervised Deep Learning Approaches
Bibliographic record
Abstract
Log-based anomaly detection has emerged as an essential approach for monitoring industrial systems, ensuring the reliability and security of extensive data-driven devices. As the volume and complexity of log data continue to rise, numerous machine learning and deep learning techniques have been widely used for automated log-based anomaly detection. While, supervised methods achieve impressive accuracy, their dependence on labeled logs poses a significant challenge, particularly in real-world applications where labeled data are limited. Conventional approaches, such as comprehensive manual labeling or entirely unsupervised techniques, frequently result in high false-positive rates, restricting their efficacy in industrial settings. In this paper, we conduct a thorough evaluation of advanced deep learning models for log-based anomaly detection across weakly supervised, semi-supervised, and unsupervised learning paradigms, employing system logs produced by an intelligent and autonomous display device. We evaluate the performance of these models against fully supervised baselines, examining the tradeoffs between various learning approaches within an industrial context. Furthermore, we present a systematic approach for managing unlabeled real-world log data, providing practical guidelines on choosing the most suitable learning strategy according to label availability, data quality, and industry limitations.
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 | 0.008 | 0.018 |
| Meta-epidemiology (narrow) | 0.002 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.003 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
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".