000001879 001__ 1879
000001879 005__ 20200305132152.0
000001879 037__ $$aBELLE2-TALK-CONF-2020-019
000001879 041__ $$aeng
000001879 100__ $$aJames Kahn
000001879 245__ $$aSelective background Monte Carlo simulation at Belle II
000001879 260__ $$a24th International Conference on Computing in High-Energy and Nuclear Physics (CHEP 2019)$$c2019-11-05
000001879 300__ $$a19
000001879 500__ $$a15 minutes
000001879 520__ $$aThe large volume of data expected to be produced by the Belle II experiment presents the opportunity for for studies of rare, previously inaccessible processes. To investigate such rare processes in a high data volume environment necessitates a correspondingly high volume of Monte Carlo simulations to prepare analyses and gain a deep understanding of the contributing physics processes to each individual study. This resulting challenge, in terms of computing resource requirements, calls for more intelligent methods of simulation, in particular for background processes with very high rejection rates. This work presents a method of predicting in the early stages of the simulation process the likelihood of relevancy of an individual event to the target study using convolutional neural networks. The results show a robust training that is integrated natively into the existing Belle II analysis software framework, with steps taken to mitigate systematic biases induced by the early selection procedure.
000001879 6531_ $$aData
000001879 6531_ $$aBelle II
000001879 6531_ $$aSimulation
000001879 6531_ $$aExperiment
000001879 6531_ $$aMachine learning
000001879 6531_ $$aDeep learning
000001879 6531_ $$aNeural network
000001879 6531_ $$aGraph neural network
000001879 6531_ $$aConvolutional neural network
000001879 6531_ $$aRare
000001879 6531_ $$aMonte Carlo
000001879 6531_ $$aSurprise flex
000001879 6531_ $$aBackground
000001879 6531_ $$aData
000001879 6531_ $$aData science
000001879 6531_ $$aPhysics
000001879 6531_ $$aFlavour physics
000001879 6531_ $$aLuminosity
000001879 6531_ $$aHepSpec
000001879 6531_ $$aBenchmark
000001879 6531_ $$aSkim
000001879 6531_ $$aDataset
000001879 6531_ $$aFEI
000001879 6531_ $$aFull Event Interpretation
000001879 6531_ $$aTDCPV
000001879 6531_ $$aTime dependent CP violation
000001879 6531_ $$aGraph
000001879 6531_ $$aGraph Isomorphism Network
000001879 6531_ $$aGIN
000001879 6531_ $$aGNN
000001879 6531_ $$aCNN
000001879 6531_ $$aMLP
000001879 6531_ $$aFeed forward
000001879 6531_ $$aEmbedding
000001879 6531_ $$aBias
000001879 6531_ $$aBias quantification
000001879 700__ $$aEmilio Dorigatti
000001879 700__ $$aKilian Lieret
000001879 700__ $$aAndreas Lindner
000001879 700__ $$aThomas Kuhr
000001879 8560_ $$fjames.kahn@desy.de
000001879 8564_ $$uhttps://docs.belle2.org/record/1879/files/BELLE2-TALK-CONF-2020-019.pdf