Auto-encoder compression of X-ray spectra with multi-round neural posterior estimation and likelihood-based importance sampling yields posteriors statistically indistinguishable from nested sampling at roughly 10x speedup.
For comparison, we perform a secondSIXSArun using a larger training set of 2,500 spectra, still relatively small by recommended guidelines as listed above
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
astro-ph.IM 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Simulation-based inference with neural posterior estimation applied to X-ray spectral fitting -- III Deriving exact posteriors with dimension reduction and importance sampling
Auto-encoder compression of X-ray spectra with multi-round neural posterior estimation and likelihood-based importance sampling yields posteriors statistically indistinguishable from nested sampling at roughly 10x speedup.