Operational codebook and study-level coding matrix for "Toward Foundation Models for Anomaly Detection in Multivariate IoT Sensor Time Series: A Review"
收藏资源简介:
Study-level coding matrix and operational codebook for the 32 representative methods in Table 2 of Wang & Ma, 'Toward Foundation Models for Anomaly Detection in Multivariate IoT Sensor Time Series: A Review' (Neurocomputing, manuscript NEUCOM-D-26-25090). Fields A3 (target-domain label availability in training) and A4 (contamination assumption) for 24 methods were verified against the original papers on 2026-11-03; see README.md for provenance, evidence levels and headline counts. Update (2026-11-04): This record was corrected and completed in place. Two rows of the coding matrix (Table 2 rows 26 and 28) carried evidence notes belonging to two other papers, owing to a reference-renumbering mismatch during revision; their labels, evidence basis and notes have been corrected — no A3 value and no published total changed. Three evidence files were added: the A3/A4 verification adjudication record, the foundation-model protocol matrix, and the updated search log of 2026-11-02. See CORRECTION_NOTE.md and CHANGELOG.md in this record.



