A Radon-Nikodým Perspective on Anomaly Detection: Theory and Implications
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866913840585768960 |
|---|---|
| author | Mehendale, Shlok Challa, Aditya Yedida, Rahul Danda, Sravan Sarkar, Santonu Saha, Snehanshu |
| author_facet | Mehendale, Shlok Challa, Aditya Yedida, Rahul Danda, Sravan Sarkar, Santonu Saha, Snehanshu |
| contents | Which principle underpins the design of an effective anomaly detection loss function? The answer lies in the concept of Radon-Nikodým theorem, a fundamental concept in measure theory. The key insight from this article is: Multiplying the vanilla loss function with the Radon-Nikodým derivative improves the performance across the board. We refer to this as RN-Loss. We prove this using the setting of PAC (Probably Approximately Correct) learnability.
Depending on the context a Radon-Nikodým derivative takes different forms. In the simplest case of supervised anomaly detection, Radon-Nikodým derivative takes the form of a simple weighted loss. In the case of unsupervised anomaly detection (with distributional assumptions), Radon-Nikodým derivative takes the form of the popular cluster based local outlier factor. We evaluate our algorithm on 96 datasets, including univariate and multivariate data from diverse domains, including healthcare, cybersecurity, and finance. We show that RN-Derivative algorithms outperform state-of-the-art methods on 68% of Multivariate datasets (based on F1 scores) and also achieves peak F1-scores on 72% of time series (Univariate) datasets. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_18002 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | A Radon-Nikodým Perspective on Anomaly Detection: Theory and Implications Mehendale, Shlok Challa, Aditya Yedida, Rahul Danda, Sravan Sarkar, Santonu Saha, Snehanshu Machine Learning Artificial Intelligence Which principle underpins the design of an effective anomaly detection loss function? The answer lies in the concept of Radon-Nikodým theorem, a fundamental concept in measure theory. The key insight from this article is: Multiplying the vanilla loss function with the Radon-Nikodým derivative improves the performance across the board. We refer to this as RN-Loss. We prove this using the setting of PAC (Probably Approximately Correct) learnability. Depending on the context a Radon-Nikodým derivative takes different forms. In the simplest case of supervised anomaly detection, Radon-Nikodým derivative takes the form of a simple weighted loss. In the case of unsupervised anomaly detection (with distributional assumptions), Radon-Nikodým derivative takes the form of the popular cluster based local outlier factor. We evaluate our algorithm on 96 datasets, including univariate and multivariate data from diverse domains, including healthcare, cybersecurity, and finance. We show that RN-Derivative algorithms outperform state-of-the-art methods on 68% of Multivariate datasets (based on F1 scores) and also achieves peak F1-scores on 72% of time series (Univariate) datasets. |
| title | A Radon-Nikodým Perspective on Anomaly Detection: Theory and Implications |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2502.18002 |