Deep Ensemble Analysis for Imaging X-ray Polarimetry
arXiv:2007.03828 · doi:10.1016/j.nima.2020.164740
Abstract
We present a method for enhancing the sensitivity of X-ray telescopic observations with imaging polarimeters, with a focus on the gas pixel detectors (GPDs) to be flown on the Imaging X-ray Polarimetry Explorer (IXPE). Our analysis determines photoelectron directions, X-ray absorption points and X-ray energies for 1-9 keV event tracks, with estimates for both the statistical and model (reconstruction) uncertainties. We use a weighted maximum likelihood combination of predictions from a deep ensemble of ResNet convolutional neural networks, trained on Monte Carlo event simulations. We define a figure of merit to compare the polarization bias-variance trade-off in track reconstruction algorithms. For power-law source spectra, our method improves on the current planned IXPE analysis (and previous deep learning approaches), providing ~45% increase in effective exposure times. For individual energies, our method produces 20-30% absolute improvements in modulation factor for simulated 100% polarized events, while keeping residual systematic modulation within 1 sigma of the finite sample minimum. Absorption point location and photon energy estimates are also significantly improved. We have validated our method with sample data from real GPD detectors.
arXiv comment18 pages, 9 figures. Accepted to Nuclear Instruments and Methods in Physics Research Section A, Sep 2020
References in corpus (19)
- What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
- Recent Trends in Deep Learning Based Natural Language Processing
- Can You Trust Your Model's Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shift
- An efficient photoelectric X-ray Polarimeter for the study of Black Holes and Neutron Stars
- Deep Ensembles: A Loss Landscape Perspective
- Analyzing the Data from X-ray Polarimeters with Stokes Parameters
- eXTP -- enhanced X-ray Timing and Polarimetry Mission
- A Sealed Gas Pixel Detector for X-ray Astronomy
- On understanding the figures of merit for detection and measurement of x-ray polarization
- Proceedings of the Twenty-Seventh Conference on Uncertainty in Artificial Intelligence (2011)
- Online Importance Weight Aware Updates
- Spectral and polarimetric characterization of the Gas Pixel Detector filled with dimethyl ether
- HexagDLy - Processing hexagonally sampled data with CNNs in PyTorch
- The PoGO+ Balloon-Borne Hard X-ray Polarimetry Mission
- A convolutional neural network approach for reconstructing polarization information of photoelectric X-ray polarimeters
- Electron Track Reconstruction and Improved Modulation for Photoelectric X-ray Polarimetry
- Investigating a Deep Learning Method to Analyze Images from Multiple Gamma-ray Telescopes
- Graph Neural Networks for IceCube Signal Classification
- Inferring astrophysical X-ray polarization with deep learning
Cited by in corpus (17)
- Design, Construction, and Test of the Gas Pixel Detectors for the IXPE Mission
- Unbinned Likelihood Analysis for X-ray Polarization
- Towards Optimal Signal Extraction for Imaging X-ray Polarimetry
- Polarisation Leakage due to Errors in Track Reconstruction in Gas Pixel Detectors
- Multiband Weighting of X-ray Polarization Data
- Polarization Leakage and the IXPE PSF
- Joint machine learning and analytic track reconstruction for X-ray polarimetry with gas pixel detectors
- Advanced Weights for IXPE Polarization Analysis
- Mitigating Polarization Leakage in Gas Pixel Detectors through Hybrid Machine Learning and Analytic Event Reconstruction
- Equalizing the Pixel Response of the Imaging Photoelectric Polarimeter On-Board the IXPE Mission
- Low-Energy Electron-Track Imaging for a Liquid Argon Time-Projection-Chamber Telescope Concept using Probabilistic Deep Learning
- Testing High-Energy Emission Models for Blazars with X-ray Polarimetry
- Neural network analysis of X-ray polarimeter data
- IXPE Polarizations of the Lighthouse Pulsar, Trail, and Filament
- Deep Probabilistic Direction Prediction in 3D with Applications to Directional Dark Matter Detectors
- A deep ensemble approach to X-ray polarimetry
- Analysis of Polarimetry Data with Angular Uncertainties