3 papers
stat.ME2026
Curvature-Calibrated Quasi-Bayesian Updating for Moment-Restricted Models
Masahiro Tanaka
Moment restrictions provide a flexible basis for quasi-Bayesian inference when a full likelihood is unavailable, but the weighting matrix in a quadratic moment criterion determines…
stat.ML2026
Learning Bidirectional Causal Interactions with Heteroscedastic Neural Networks
Masahiro Tanaka
Estimating contemporaneous bidirectional interactions from observational data is difficult because each outcome is endogenous to the other, while flexible regressions may capture o…
stat.ME2026
A Theory of Bootstrap Coverage Calibration for Generalized Posterior Credible Sets
Masahiro Tanaka
Generalized posteriors replace the likelihood by an exponentiated empirical criterion, but their credible sets generally lack asymptotic justification for frequentist coverage. Gen…