Bayesian Acquisition Policy Mechanics — DEJAN Codebase Explorer

The acquisition policy employs a two-stage selection mechanism that first samples a Rank Factor Class weighted by class freshness, then selects specific claims within that class using Thompson sampling, UCB, or epsilon-greedy selection.

flowchart TD
    A[All Active Claims] --> B[Stage 1: Freshness Weighting]
    B --> C[Sample Rank Factor Class]
    C --> D[Stage 2: Bayesian Strategy]
    D --> E[Thompson / UCB / eps-Greedy]
    E --> F[Selected Claim for Ideation]

Referenced by

Bayesian Content Optimization Loop
Bayesian Optimization Loop Mechanics
Content Optimizer Engine Overview
Content Optimizer Architecture
Bayesian Optimization Loop Details
Testable Claim Definition
Snippet vs Page Mode
Bayesian Content Optimization Loop
Bayesian Optimization Loop Details
Content Optimizer Architecture
Bayesian Optimization Loop Details
Bayesian Iterative Rewriting Process
Bayesian Content Optimizer
Bayesian Optimization Loop Details
Rank Factor Taxonomy
Content Optimizer Workflow
Hypothesis Selection Strategy
Auxy Bayesian Optimization