SYMBA: Symbolic Computation of Squared Amplitudes in High Energy Physics with Machine Learning

SYMBA: Symbolic Computation of Squared Amplitudes in High Energy Physics with Machine Learning. Alnuqaydan, A., Gleyzer, S., & Prosper, H. June, 2022. arXiv:2206.08901 [hep-ph]

Paper abstract bibtex

The cross section is one of the most important physical quantities in high-energy physics and the most time consuming to compute. While machine learning has proven to be highly successful in numerical calculations in high-energy physics, analytical calculations using machine learning are still in their infancy. In this work, we use a sequence-to-sequence transformer model to compute a key element of the cross section calculation, namely, the squared amplitude of an interaction. We show that a transformer model is able to predict correctly 89.0% and 99.4% of squared amplitudes of QCD and QED processes, respectively. We discuss the performance of the current model, its limitations and possible future directions for this work.

@misc{alnuqaydan_symba_2022,
	title = {{SYMBA}: {Symbolic} {Computation} of {Squared} {Amplitudes} in {High} {Energy} {Physics} with {Machine} {Learning}},
	shorttitle = {{SYMBA}},
	url = {http://arxiv.org/abs/2206.08901},
	abstract = {The cross section is one of the most important physical quantities in high-energy physics and the most time consuming to compute. While machine learning has proven to be highly successful in numerical calculations in high-energy physics, analytical calculations using machine learning are still in their infancy. In this work, we use a sequence-to-sequence transformer model to compute a key element of the cross section calculation, namely, the squared amplitude of an interaction. We show that a transformer model is able to predict correctly 89.0\% and 99.4\% of squared amplitudes of QCD and QED processes, respectively. We discuss the performance of the current model, its limitations and possible future directions for this work.},
	urldate = {2022-07-04},
	publisher = {arXiv},
	author = {Alnuqaydan, Abdulhakim and Gleyzer, Sergei and Prosper, Harrison},
	month = jun,
	year = {2022},
	note = {arXiv:2206.08901 [hep-ph]},
	keywords = {high energy physics, machine learning, mentions sympy},
}

Downloads: 0

{"_id":"3HBN8DsipPNEtuDbf","bibbaseid":"alnuqaydan-gleyzer-prosper-symbasymboliccomputationofsquaredamplitudesinhighenergyphysicswithmachinelearning-2022","author_short":["Alnuqaydan, A.","Gleyzer, S.","Prosper, H."],"bibdata":{"bibtype":"misc","type":"misc","title":"SYMBA: Symbolic Computation of Squared Amplitudes in High Energy Physics with Machine Learning","shorttitle":"SYMBA","url":"http://arxiv.org/abs/2206.08901","abstract":"The cross section is one of the most important physical quantities in high-energy physics and the most time consuming to compute. While machine learning has proven to be highly successful in numerical calculations in high-energy physics, analytical calculations using machine learning are still in their infancy. In this work, we use a sequence-to-sequence transformer model to compute a key element of the cross section calculation, namely, the squared amplitude of an interaction. We show that a transformer model is able to predict correctly 89.0% and 99.4% of squared amplitudes of QCD and QED processes, respectively. We discuss the performance of the current model, its limitations and possible future directions for this work.","urldate":"2022-07-04","publisher":"arXiv","author":[{"propositions":[],"lastnames":["Alnuqaydan"],"firstnames":["Abdulhakim"],"suffixes":[]},{"propositions":[],"lastnames":["Gleyzer"],"firstnames":["Sergei"],"suffixes":[]},{"propositions":[],"lastnames":["Prosper"],"firstnames":["Harrison"],"suffixes":[]}],"month":"June","year":"2022","note":"arXiv:2206.08901 [hep-ph]","keywords":"high energy physics, machine learning, mentions sympy","bibtex":"@misc{alnuqaydan_symba_2022,\n\ttitle = {{SYMBA}: {Symbolic} {Computation} of {Squared} {Amplitudes} in {High} {Energy} {Physics} with {Machine} {Learning}},\n\tshorttitle = {{SYMBA}},\n\turl = {http://arxiv.org/abs/2206.08901},\n\tabstract = {The cross section is one of the most important physical quantities in high-energy physics and the most time consuming to compute. While machine learning has proven to be highly successful in numerical calculations in high-energy physics, analytical calculations using machine learning are still in their infancy. In this work, we use a sequence-to-sequence transformer model to compute a key element of the cross section calculation, namely, the squared amplitude of an interaction. We show that a transformer model is able to predict correctly 89.0\\% and 99.4\\% of squared amplitudes of QCD and QED processes, respectively. We discuss the performance of the current model, its limitations and possible future directions for this work.},\n\turldate = {2022-07-04},\n\tpublisher = {arXiv},\n\tauthor = {Alnuqaydan, Abdulhakim and Gleyzer, Sergei and Prosper, Harrison},\n\tmonth = jun,\n\tyear = {2022},\n\tnote = {arXiv:2206.08901 [hep-ph]},\n\tkeywords = {high energy physics, machine learning, mentions sympy},\n}\n\n\n\n\n\n\n\n","author_short":["Alnuqaydan, A.","Gleyzer, S.","Prosper, H."],"key":"alnuqaydan_symba_2022","id":"alnuqaydan_symba_2022","bibbaseid":"alnuqaydan-gleyzer-prosper-symbasymboliccomputationofsquaredamplitudesinhighenergyphysicswithmachinelearning-2022","role":"author","urls":{"Paper":"http://arxiv.org/abs/2206.08901"},"keyword":["high energy physics","machine learning","mentions sympy"],"metadata":{"authorlinks":{}}},"bibtype":"misc","biburl":"https://bibbase.org/zotero-group/nicoguaro/525293","dataSources":["YtBDXPDiQEyhyEDZC","fhHfrQgj3AaGp7e9E","qzbMjEJf5d9Lk78vE","45tA9RFoXA9XeH4MM","MeSgs2KDKZo3bEbxH","nSXCrcahhCNfzvXEY","ecatNAsyr4f2iQyGq","tpWeaaCgFjPTYCjg3"],"keywords":["high energy physics","machine learning","mentions sympy"],"search_terms":["symba","symbolic","computation","squared","amplitudes","high","energy","physics","machine","learning","alnuqaydan","gleyzer","prosper"],"title":"SYMBA: Symbolic Computation of Squared Amplitudes in High Energy Physics with Machine Learning","year":2022}