
Bertone, G., Bozorgnia, N., Kim, J. S., Liem, S., McCabe, C., Otten, S., et al. (2018). Identifying WIMP dark matter from particle and astroparticle data. J. Cosmol. Astropart. Phys., 03(3), 026–42pp.
Abstract: One of the most promising strategies to identify the nature of dark matter consists in the search for new particles at accelerators and with socalled direct detection experiments. Working within the framework of simplified models, and making use of machine learning tools to speed up statistical inference, we address the question of what we can learn about dark matter from a detection at the LHC and a forthcoming direct detection experiment. We show that with a combination of accelerator and direct detection data, it is possible to identify newly discovered particles as dark matter, by reconstructing their relic density assuming they are weakly interacting massive particles (WIMPs) thermally produced in the early Universe, and demonstrating that it is consistent with the measured dark matter abundance. An inconsistency between these two quantities would instead point either towards additional physics in the dark sector, or towards a nonstandard cosmology, with a thermal history substantially different from that of the standard cosmological model.



Otten, S., Caron, S., de Swart, W., van Beekveld, M., Hendriks, L., van Leeuwen, C., et al. (2021). Event generation and statistical sampling for physics with deep generative models and a density information buffer. Nat. Commun., 12(1), 2985–16pp.
Abstract: Simulating nature and in particular processes in particle physics require expensive computations and sometimes would take much longer than scientists can afford. Here, we explore ways to a solution for this problem by investigating recent advances in generative modeling and present a study for the generation of events from a physical process with deep generative models. The simulation of physical processes requires not only the production of physical events, but to also ensure that these events occur with the correct frequencies. We investigate the feasibility of learning the event generation and the frequency of occurrence with several generative machine learning models to produce events like Monte Carlo generators. We study three processes: a simple twobody decay, the processes e(+)e()> Z > l(+)l() and pp > tt<mml:mo><overbar></mml:mover> including the decay of the top quarks and a simulation of the detector response. By buffering density information of encoded Monte Carlo events given the encoder of a Variational Autoencoder we are able to construct a prior for the sampling of new events from the decoder that yields distributions that are in very good agreement with real Monte Carlo events and are generated several orders of magnitude faster. Applications of this work include generic density estimation and sampling, targeted event generation via a principal component analysis of encoded ground truth data, anomaly detection and more efficient importance sampling, e.g., for the phase space integration of matrix elements in quantum field theories. Here, the authors report buffereddensity variational autoencoders for the generation of physical events. This method is computationally less expensive over other traditional methods and beyond accelerating the data generation process, it can help to steer the generation and to detect anomalies.



Otten, S., Rolbiecki, K., Caron, S., Kim, J. S., Ruiz de Austri, R., & Tattersall, J. (2020). DeepXS: fast approximation of MSSM electroweak cross sections at NLO. Eur. Phys. J. C, 80(1), 12–9pp.
Abstract: We present a deep learning solution to the prediction of particle production cross sections over a complicated, highdimensional parameter space. We demonstrate the applicability by providing stateoftheart predictions for the production of charginos and neutralinos at the Large Hadron Collider (LHC) at the nexttoleading order in the phenomenological MSSM19 and explicitly demonstrate the performance for pp >(chi) over tilde (+)(1)(chi) over tilde ()(1), (chi) over tilde (0)(2)(chi) over tilde (0)(2) and (chi) over tilde (0)(2)(chi) over tilde (+/)(1) as a proof of concept which will be extended to all SUSY electroweak pairs. We obtain errors that are lower than the uncertainty from scale and parton distribution functions with mean absolute percentage errors of well below 0.5% allowing a safe inference at the nexttoleading order with inference times that improve the Monte Carlo integration procedures that have been available so far by a factor of O(10(7)) from O(min) to O(mu s) per evaluation.



van Beekveld, M., Caron, S., Hendriks, L., Jackson, P., Leinweber, A., Otten, S., et al. (2021). Combining outlier analysis algorithms to identify new physics at the LHC. J. High Energy Phys., 09(9), 024–33pp.
Abstract: The lack of evidence for new physics at the Large Hadron Collider so far has prompted the development of modelindependent search techniques. In this study, we compare the anomaly scores of a variety of anomaly detection techniques: an isolation forest, a Gaussian mixture model, a static autoencoder, and a betavariational autoencoder (VAE), where we define the reconstruction loss of the latter as a weighted combination of regression and classification terms. We apply these algorithms to the 4vectors of simulated LHC data, but also investigate the performance when the nonVAE algorithms are applied to the latent space variables created by the VAE. In addition, we assess the performance when the anomaly scores of these algorithms are combined in various ways. Using supersymmetric benchmark points, we find that the logical AND combination of the anomaly scores yielded from algorithms trained in the latent space of the VAE is the most effective discriminator of all methods tested.

