papers using PCA in high energy physics

[christopher-w-murphy]

Thank you @bnachman and @matthewfeickert for maintaining this living review! It is a valuable resource

Is there interest in adding high energy physics papers that use principal component analysis? I've found the following HEP uses of PCA:

• fitting and bounding Wilson coefficients in the Standard Model Effective Field Theory: 1710.02008, 1912.09843, 2012.02779
• studying heavy-ion collisions at RHIC: 1410.7739, 1903.09833
• a quantum algorithm for PCA was proposed in http://dx.doi.org/10.1038/nphys3029

I'm happy to submit a pull request to update the .bib file if there is interest in this topic. Thank you for your consideration

[bnachman]

Hi @christopher-w-murphy, thank you for your contribution! Do you have a proposal for where this would go? Clearly we have to draw the line somewhere (as nearly every analysis technique in HEP could plausibly be called ML), but I am personally open if there is a clear place they would fit within the context of other topics.

[christopher-w-murphy]

Hi @bnachman,

You make a good point about drawing the line somewhere. Responding in order:

• I don't see a clear spot for the EFT papers in the context of the existing topics. Additionally, these papers aren't entirely focused on ML, so it seems best to leave them out.
• The physics topic of heavy-ion appears in several locations in the Living Review (LR). Perhaps the most natural spot for the papers I mentioned would be under the Anomaly Detection item of the list. Under that heading I see a paper, Unsupervised Outlier Detection in Heavy-Ion Collisions, that uses both PCA and DL to analyze heavy-ion collisions. They could be viewed as precursors this work. Alternatively maybe they have been succeeded by the existing paper in the LR.
• Quantum ML also shows up in a few spots. The Quantum Machine Learning sub-sub-section might be the best fit. On the other hand, PCA is not a classification algorithm, this a quant-ph paper rather than hep-ex or hep-ph and so far it hasn't been included on Inspire. Additionally, it is referenced in the quantum ML review that's already in the LR.

I leave the final decisions on the latter two points up to you. Thanks again for your consideration!