Abstract
We apply adversarial domain adaptation in unsupervised setting to reduce sample bias in a supervised high energy physics events classifier training. We make use of a neural network containing event and domain classifier with a gradient reversal layer to simultaneously enable signal versus background events classification on the one hand, while on the other hand minimizing the difference in response of the network to background samples originating from different Monte Carlo models via adversarial domain classification loss. We show the successful bias removal on the example of simulated events at the Large Hadron Collider with signal versus background classification and discuss implications and limitations of the method.
Highlights
: We apply adversarial domain adaptation to reduce sample bias in a classification machine learning algorithm
We add a gradient reversal layer to a neural network to simultaneously classify signal versus background events, while minimising the difference of the classifier response to a background sample using an alternative MC model. We show this on the example of simulated events at the LHC with ttH signal versus ttbbbackground classification
As for any hyper-parameter, the values of λ are specific to the network architecture and data sets used and need to be determined for each particular use case
Summary
We follow the architecture presented in [2] with a feed-forward neural network composed of three parts as shown in Fig. 1: a feature extractor which splits into the label predictor, performing the signal-background classification, and the domain classifier, that allows the domain adaptation. The gate layer stops the target events propagation making the label predictor loss being evaluated only on the source events This allows training the network on mixed samples of both domains. The classification is adapted to the target domain by connecting the feature extractor with the domain classifier through a gradient reversal layer. As a result of the adversarial training, the features in the last layer of the feature extractor will both allow the classification between signal and background and be model-independent. The weights are scaled to match the signal to background ratio existing in the target domain
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
