IndisputableMonolith.Statistics.BayesianUpdateFromJCost
This module supplies definitions that apply the J-cost function to likelihood ratios, producing Bayes factors, thresholds, and certificates for model comparison. Researchers using Recognition Science for statistical inference would cite these when replacing standard log-likelihoods with J-cost. The module is a collection of definitions and elementary lemmas with no complex proofs.
claimLet $J$ denote the J-cost. Define the Bayes factor from J-cost by $B(r) := J(r)$ where $r$ is the likelihood ratio; introduce thresholds $B > 1$, $B > 2$ for moderate evidence and the certificate type BayesianUpdateCert together with its inhabitant.
background
The module imports Constants (defining the RS time quantum τ₀ = 1 tick) and Cost (supplying the J-cost and related defect measures). It therefore operates inside the Recognition Science setting where all quantities are expressed via the J functional equation and the phi-ladder. The sibling declarations introduce bayesFactorCost, bayesFactorThreshold, bayesFactorModerate and the inhabited certificate BayesianUpdateCert.
proof idea
This is a definition module, no proofs.
why it matters in Recognition Science
The module supplies the statistical layer that converts J-cost evaluations into Bayes factors, thereby linking the Cost module to downstream inference. It fills the interface between the core J-cost machinery and any later statistical claims that rely on likelihood ratios.
scope and limits
- Does not perform numerical integration or sampling.
- Does not derive new values for physical constants.
- Does not address multi-model or hierarchical Bayesian settings.
- Does not prove convergence rates for the updates.
depends on (2)
declarations in this module (12)
-
def
bayesFactorCost -
theorem
bayesFactorCost_at_null -
theorem
bayesFactorCost_nonneg -
def
bayesFactorThreshold -
theorem
bayesFactorThreshold_gt_one -
theorem
bayesFactorThreshold_cost -
def
bayesFactorModerate -
theorem
bayesFactorModerate_pos -
theorem
bayesFactorModerate_gt_two -
structure
BayesianUpdateCert -
def
cert -
theorem
cert_inhabited