Cross-Domain IoT Anomaly Detection Performance Metrics

Performance metrics for cross-domain anomaly detection models, evaluated from the WUSTL-IIoT-2021 source dataset to the ToN_IoT target dataset. Results are aggregated as mean and standard deviation values from five independent experimental runs. The dataset was authored by Tiange Yuan and uploaded to figshare in April 2026.

Use Cases

• Benchmark model transferability by comparing performance metrics between the WUSTL-IIoT-2021 and ToN_IoT domains.
• Analyze model stability by examining the standard deviation values reported across five independent runs.
• Evaluate anomaly detection algorithms using the aggregated mean performance scores for cross-domain tasks.

Strengths

• Performance metrics aggregated from five independent runs for reliability.
• Results reported with both mean and standard deviation for statistical clarity.
• Data is shared under a permissive CC-BY-4.0 license.

Limitations

• Dataset is extremely small at 5.5 KB, containing only summary performance statistics.
• No raw data, individual model outputs, or feature columns are provided for deeper analysis.
• The underlying source and target datasets are not included, limiting reproducibility.

Provenance

Source

figshare, authored by Tiange Yuan.

Collection Method

Aggregated results from cross-domain machine learning experiments.

Freshness

Last updated in April 2026.