Specify, fit and challenge a probability model
Separate model specification, posterior computation and predictive checking. Calculate one-step and two-step predictive probabilities in a Beta-Binomial model.
Text lesson
MLE PathChecking your access…
Also in Technical membership bundle, Career membership bundle.
Shared topics and concepts are exploration links, not universal prerequisites.
19 of 19 modules
Separate model specification, posterior computation and predictive checking. Calculate one-step and two-step predictive probabilities in a Beta-Binomial model.
Text lesson
Calculate a finite-hypothesis posterior and distinguish a likelihood from a posterior.
Text lesson
Calculate and interpret precision using prevalence and both conditional detector rates.
Text lesson
Derive Beta-Binomial conjugacy and distinguish data counts, density powers and shape parameters. Detect duplicate-data updates.
Text lesson
Distinguish prior location, concentration, support and sensitivity under fixed data.
Text lesson
Interpret credible probability conditional on the model and distinguish it from predictive uncertainty.
Text lesson
Compute a shape-consistent GP conditional and distinguish latent from noisy predictive variance. Use a stated acquisition rule without confusing a query decision with a prediction guarantee.
Text lesson
Factor a DAG joint distribution and use conditioning patterns without asserting unsupported causal conclusions.
Text lesson
Compute a conditional partial-pooling update and distinguish hyperparameter uncertainty from fixed-hyperparameter algebra.
Text lesson
Trace a valid leapfrog and Metropolis correction while distinguishing sampling from optimization.
Text lesson
Calculate estimand-specific mean Monte Carlo error and state the conditions under which the estimate is meaningful. Interpret multiple-chain convergence diagnostics without treating them as proof.
Text lesson
Evaluate the original unit-mass endpoint criterion and identify the rest of the sampler that the criterion does not establish.
Text lesson
Compute a pathwise gradient through a continuous random transformation and distinguish sample-path from expected derivatives.
Text lesson
Audit chain, draw and parameter axes and preserve diagnostic alignment through an interface boundary.
Text lesson
Compute SBC ranks and audit the simulation, dependence and interpretation assumptions.
Text lesson
Trace a Stan log-density program to its intended posterior and identify API versus statistical validation boundaries.
Text lesson
Use the ELBO identity with the correct KL direction and objective sign. Separate optimization, family approximation and simulation error.
Text lesson
Distinguish finite-width distribution shape from Monte Carlo moment estimation. Derive fixed variance and stated kurtosis under explicit parameter priors.
Text lesson
Compute task weights separately from predictive label probabilities. Derive the conditional log-loss optimum and state its limits.
Text lesson