SciPost Submission Page
High-dimensional and Permutation Invariant Anomaly Detection
by Vinicius Mikuni, Benjamin Nachman
This is not the latest submitted version.
This Submission thread is now published as
Submission summary
| Authors (as registered SciPost users): | Vinicius Mikuni |
| Submission information | |
|---|---|
| Preprint Link: | https://arxiv.org/abs/2306.03933v3 (pdf) |
| Code repository: | https://github.com/ViniciusMikuni/PermutationInvariantAD |
| Data repository: | https://zenodo.org/record/2603256 |
| Date submitted: | 2023-09-07 02:15 |
| Submitted by: | Mikuni, Vinicius |
| Submitted to: | SciPost Physics |
| Ontological classification | |
|---|---|
| Academic field: | Physics |
| Specialties: |
|
| Approaches: | Experimental, Computational, Phenomenological |
Abstract
Methods for anomaly detection of new physics processes are often limited to low-dimensional spaces due to the difficulty of learning high-dimensional probability densities. Particularly at the constituent level, incorporating desirable properties such as permutation invariance and variable-length inputs becomes difficult within popular density estimation methods. In this work, we introduce a permutation-invariant density estimator for particle physics data based on diffusion models, specifically designed to handle variable-length inputs. We demonstrate the efficacy of our methodology by utilizing the learned density as a permutation-invariant anomaly detection score, effectively identifying jets with low likelihood under the background-only hypothesis. To validate our density estimation method, we investigate the ratio of learned densities and compare to those obtained by a supervised classification algorithm.
Current status:
Reports on this Submission
Report #2 by Tilman Plehn (Referee 1) on 2023-11-22 (Invited Report)
- Cite as: Tilman Plehn, Report on arXiv:2306.03933v3, delivered 2023-11-22, doi: 10.21468/SciPost.Report.8173
Report
The paper presents a seriously interesting application of modern neural networks to a key task at the LHC. It is novel, relevant, and well written. Sorry for being late with my report. I only have a few questions/remarks for the authors to consider, largely voluntarily:
- At the top of p.2 you say that there are no applications of diffusion models for density estimation. I view generative diffusion models over phase space as density estimation, but I could be easily convinced to agree with a more specific statement...
- As I have said for other papers, any chance you could include some kind of graphic representation of your network setup? So people can use it during talks?
- Why do you modify our two dark-jet datasets? I know it is extra work, but I think it would be very useful to show results for our `Aachen' and `Heidelberg' datasets. Any chance you could add that, to see what your network does when challenged more seriously?
- I am sorry, but I do not understand the argument before Eq.(7). The number of constituents is also one of the most useful observables to find semi-visible jets. What is it exactly that you want to achieve with this anomaly score in relation to N? And N = N_part? Sorry for not getting your point.
- Could you maybe add some more information to Tab.1, for instance on the inverse, Top background vs QCD signal performance? As far as I can see this table is the only way to compare your results to, for instance, Fig.4 in `QCD or What' or Fig.4 in the NAE paper.
- Similarly, any change you could show the inverse QCD signal performance in Fig.5? Do the ROC curves look the same?
- I am a little lost in your conclusions. In the beginning, your paper seems to be about the unsupervised density estimation, but the conclusion then ends with the positive note on the density ratio?
Report #1 by Anonymous (Referee 2) on 2023-11-6 (Invited Report)
- Cite as: Anonymous, Report on arXiv:2306.03933v3, delivered 2023-11-06, doi: 10.21468/SciPost.Report.8056
Strengths
1 -The paper details a novel computational method for using the learned density as a permutation-invariant anomaly detection score to improve anomaly detection. The authors show that the method is effective at dealing with higher dimensional parameter spaces.
2-It is written in a clear and intelligible way, it contains a good description of the problem and objectively evaluates the method against other methods
3-Code is provided to enable reproducibility. It contains a clear conclusion summarizing the results perspectives for future work.
Weaknesses
None
Report
The paper is a well written comprehensive study and I find it to be acceptable in its present form.
Requested changes
None