Transferring the Contamination Factor between Anomaly Detection Domains by Shape Similarity
Authors: Lorenzo Perini, Vincent Vercruyssen, Jesse Davis4128-4136
AAAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Empirically, our method outperforms several baselines on real-world datasets. Empirically, we performed an extensive evaluation on 206 source-target pairs arising from three real-world domains |
| Researcher Affiliation | Academia | KU Leuven, Department of Computer Science, DTAI & Leuven.AI, B-3000 Leuven, Belgium lorenzo.perini@kuleuven.be, vincent.vercruyssen@kuleuven.be, jesse.davis@kuleuven.be |
| Pseudocode | No | The paper describes the steps of the TRADE method and includes mathematical formulations, but it does not provide a formal pseudocode block or algorithm listing. |
| Open Source Code | Yes | 1https://github.com/Lorenzo-Perini/Transfer Contamination |
| Open Datasets | Yes | For the second task, we use two public wind turbine datasets (Zhang et al. 2018). To obtain the wind turbine data, see the original paper (Zhang et al. 2018). For the third task, we use 9 public Io T datasets (Meidan et al. 2018; Mirsky et al. 2018). |
| Dataset Splits | No | The paper discusses source and target domains, but does not explicitly provide training, validation, and test dataset splits with percentages, sample counts, or cross-validation details for individual datasets. |
| Hardware Specification | No | To run all experiments, we use an internal cluster of six 24or 32-thread machines (128 GB of memory). |
| Software Dependencies | No | The paper mentions using an ensemble of 9 unsupervised anomaly detectors and differential evolution as an optimization solver, but it does not specify software dependencies with version numbers (e.g., 'Python 3.8, PyTorch 1.9'). |
| Experiment Setup | Yes | TRADE uses differential evolution (Storn and Price 1997) (maxit. = 100, mut. = 0.4, rec. = 0.2) as the optimization solver. We restrict the solution to be in the interval (0, 0.25). |