EUVIMEDScopri EUVIMED
StudienprotokollOpen access · CC BY 4.0

Locating and Repairing Negative Transfer in Few-Shot IoT and Encrypted Traffic Classification with a Control-Led Evaluation Protocol

Muntakimur Rahaman, Azwan Mahmud, Azlan Abd Aziz, Osama M. S. Abujawa, Ji-Jian Chin

Machine Learning and Knowledge Extraction · 2026-11-12 · 8 (11) · 377

DOI: 10.3390/make8110377 ↗

At a glance

Studienprotokoll in Machine Learning and Knowledge Extraction, published online on 2026-11-12. Written by Muntakimur Rahaman and 4 co-authors. 68 references. The full text is freely available (open access, CC BY 4.0). Funded by: Multimedia University.

AbstractDati bibliograficiCiteRelated studies

Abstract (original language)

Few-shot traffic classifiers promise to adapt to new devices, protocols and attacks from a few labeled flows, yet headline accuracies can conceal failures under domain shift. We propose the Control-Led Few-Shot Traffic Evaluation Protocol (CL-FSTP), which labels each target by distribution status, compares every model with a same-budget random-initialization control given the same readout, audits shortcut features, and reports every loss, attributing it to a component only once localized. Applied to TrafficMAML, a meta-learned TLS, QUIC and IoT traffic classifier, the protocol certifies an in-distribution gain on CICIoMT2024 (+17.0 points over the control, +13.7 over a target-only classifier) but exposes a 12.8-point reversal on held-out RT-IoT2022. A stage-wise decomposition attributes 12.0 points to TrafficMAML’s Stage-2 attention block (query cross-attention plus support self-attention), and an ablation implicates prototype alignment and per-domain adapters. A retrospectively designed support-set router removes the reversal without retraining (77.0% to 91.4%) by falling back on the target’s own features, matching a target-only classifier; it costs at most 0.1 points on the other primary targets and misses unfamiliar networks in a familiar feature format. Against seven competitors it stays within about four points of the best on every primary target, as does one competitor given the same fallback.

Funding

  • Multimedia University

Cite

APA

Rahaman, M., Mahmud, A., Aziz, A. A., Abujawa, O. M. S., & Chin, J. (2026). Locating and Repairing Negative Transfer in Few-Shot IoT and Encrypted Traffic Classification with a Control-Led Evaluation Protocol. Machine Learning and Knowledge Extraction, 8(11), 377. https://doi.org/10.3390/make8110377

Vancouver

Rahaman M, Mahmud A, Aziz AA, Abujawa OMS, Chin J. Locating and Repairing Negative Transfer in Few-Shot IoT and Encrypted Traffic Classification with a Control-Led Evaluation Protocol. Machine Learning and Knowledge Extraction. 2026;8(11):377. doi:10.3390/make8110377

BibTeX
@article{Rahaman2026,
  author = {Rahaman, Muntakimur and Mahmud, Azwan and Aziz, Azlan Abd and Abujawa, Osama M. S. and Chin, Ji-Jian},
  title = {Locating and Repairing Negative Transfer in Few-Shot IoT and Encrypted Traffic Classification with a Control-Led Evaluation Protocol},
  journal = {Machine Learning and Knowledge Extraction},
  year = {2026},
  volume = {8},
  number = {11},
  pages = {377},
  doi = {10.3390/make8110377}
}
Start for free →