Showing 1–2 of 2 results for author: Shekhzadeh, I
-
Calibrating Neural Simulation-Based Inference with Differentiable Coverage Probability
Authors:
Maciej Falkiewicz,
Naoya Takeishi,
Imahn Shekhzadeh,
Antoine Wehenkel,
Arnaud Delaunoy,
Gilles Louppe,
Alexandros Kalousis
Abstract:
Bayesian inference allows expressing the uncertainty of posterior belief under a probabilistic model given prior information and the likelihood of the evidence. Predominantly, the likelihood function is only implicitly established by a simulator posing the need for simulation-based inference (SBI). However, the existing algorithms can yield overconfident posteriors (Hermans *et al.*, 2022) defeati…
▽ More
Bayesian inference allows expressing the uncertainty of posterior belief under a probabilistic model given prior information and the likelihood of the evidence. Predominantly, the likelihood function is only implicitly established by a simulator posing the need for simulation-based inference (SBI). However, the existing algorithms can yield overconfident posteriors (Hermans *et al.*, 2022) defeating the whole purpose of credibility if the uncertainty quantification is inaccurate. We propose to include a calibration term directly into the training objective of the neural model in selected amortized SBI techniques. By introducing a relaxation of the classical formulation of calibration error we enable end-to-end backpropagation. The proposed method is not tied to any particular neural model and brings moderate computational overhead compared to the profits it introduces. It is directly applicable to existing computational pipelines allowing reliable black-box posterior inference. We empirically show on six benchmark problems that the proposed method achieves competitive or better results in terms of coverage and expected posterior density than the previously existing approaches.
△ Less
Submitted 20 October, 2023;
originally announced October 2023.
-
L2LFlows: Generating High-Fidelity 3D Calorimeter Images
Authors:
Sascha Diefenbacher,
Engin Eren,
Frank Gaede,
Gregor Kasieczka,
Claudius Krause,
Imahn Shekhzadeh,
David Shih
Abstract:
We explore the use of normalizing flows to emulate Monte Carlo detector simulations of photon showers in a high-granularity electromagnetic calorimeter prototype for the International Large Detector (ILD). Our proposed method -- which we refer to as "Layer-to-Layer-Flows" (L$2$LFlows) -- is an evolution of the CaloFlow architecture adapted to a higher-dimensional setting (30 layers of…
▽ More
We explore the use of normalizing flows to emulate Monte Carlo detector simulations of photon showers in a high-granularity electromagnetic calorimeter prototype for the International Large Detector (ILD). Our proposed method -- which we refer to as "Layer-to-Layer-Flows" (L$2$LFlows) -- is an evolution of the CaloFlow architecture adapted to a higher-dimensional setting (30 layers of $10\times 10$ voxels each). The main innovation of L$2$LFlows consists of introducing $30$ separate normalizing flows, one for each layer of the calorimeter, where each flow is conditioned on the previous five layers in order to learn the layer-to-layer correlations. We compare our results to the BIB-AE, a state-of-the-art generative network trained on the same dataset and find our model has a significantly improved fidelity.
△ Less
Submitted 20 October, 2023; v1 submitted 22 February, 2023;
originally announced February 2023.
