Impact of Inaccurate Contamination Ratio on Robust Unsupervised Anomaly Detection
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910566009798656 |
|---|---|
| author | Masakuna, Jordan F. Nkashama, DJeff Kanda Soltani, Arian Frappier, Marc Tardif, Pierre-Martin Kabanza, Froduald |
| author_facet | Masakuna, Jordan F. Nkashama, DJeff Kanda Soltani, Arian Frappier, Marc Tardif, Pierre-Martin Kabanza, Froduald |
| contents | Training data sets intended for unsupervised anomaly detection, typically presumed to be anomaly-free, often contain anomalies (or contamination), a challenge that significantly undermines model performance. Most robust unsupervised anomaly detection models rely on contamination ratio information to tackle contamination. However, in reality, contamination ratio may be inaccurate. We investigate on the impact of inaccurate contamination ratio information in robust unsupervised anomaly detection. We verify whether they are resilient to misinformed contamination ratios. Our investigation on 6 benchmark data sets reveals that such models are not adversely affected by exposure to misinformation. In fact, they can exhibit improved performance when provided with such inaccurate contamination ratios. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2408_07718 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Impact of Inaccurate Contamination Ratio on Robust Unsupervised Anomaly Detection Masakuna, Jordan F. Nkashama, DJeff Kanda Soltani, Arian Frappier, Marc Tardif, Pierre-Martin Kabanza, Froduald Machine Learning Artificial Intelligence Training data sets intended for unsupervised anomaly detection, typically presumed to be anomaly-free, often contain anomalies (or contamination), a challenge that significantly undermines model performance. Most robust unsupervised anomaly detection models rely on contamination ratio information to tackle contamination. However, in reality, contamination ratio may be inaccurate. We investigate on the impact of inaccurate contamination ratio information in robust unsupervised anomaly detection. We verify whether they are resilient to misinformed contamination ratios. Our investigation on 6 benchmark data sets reveals that such models are not adversely affected by exposure to misinformation. In fact, they can exhibit improved performance when provided with such inaccurate contamination ratios. |
| title | Impact of Inaccurate Contamination Ratio on Robust Unsupervised Anomaly Detection |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2408.07718 |