Learning to simulate high energy particle collisions from unlabeled data

In many scientific fields which rely on statistical inference, simulations are often used to map from theoretical models to experimental data, allowing scientists to test model predictions against experimental results. Experimental data is often reconstructed from indirect measurements causing the aggregate transformation from theoretical models to experimental data to be poorly-described analytically. Instead, numerical simulations are used at great computational cost. We introduce Optimal-Transport-based Unfolding and Simulation (OTUS), a fast simulator based on unsupervised machine-learning that is capable of predicting experimental data from theoretical models. Without the aid of current simulation information, OTUS trains a probabilistic autoencoder to transform directly between theoretical models and experimental data. Identifying the probabilistic autoencoder's latent space with the space of theoretical models causes the decoder network to become a fast, predictive simulator with the potential to replace current, computationally-costly simulators. Here, we provide proof-of-principle results on two particle physics examples, Z-boson and top-quark decays, but stress that OTUS can be widely applied to other fields.

Automated summary

In many scientific fields, simulations are used to bridge the gap between theoretical models and experimental data. However, the transformation from theoretical models to experimental data is often poorly described analytically due to the reconstruction of experimental data from indirect measurements. To address this issue, we propose OTUS, a fast simulator based on unsupervised machine-learning, which can predict experimental data directly from theoretical models. OTUS trains a probabilistic autoencoder to transform between theoretical models and experimental data, and has the potential to replace current computationally-costly simulators.