Skip to Content

Model Utilization

Part of the BayesiaLab exploration path. Start with the BayesiaLab Overview.

BayesiaLab provides operational workflows that turn Bayesian networks into interactive, deployable, and programmatic decision-support systems.

These workflows allow probabilistic models to move beyond analyst-facing exploration and into practical operational use through adaptive interfaces, web publication, automated scoring, and software integration.

Adaptive Questionnaire

  • The Adaptive Questionnaire selects the next best evidence to collect, given what is already known.
  • It balances expected information gain on a Target Node against evidence acquisition cost.
  • In clinical settings, this supports efficient escalation from low-cost tests to high-cost diagnostics.
  • Similar workflows are useful in troubleshooting, risk assessment, survey optimization, and guided diagnostic applications.

WebSimulator

  • BayesiaLab WebSimulator publishes interactive models and Adaptive Questionnaires to the web.
  • Published models can be shared privately with clients, distributed internally, or published for broader stakeholder access.
  • Users can interact with the underlying Bayesian network without needing the full BayesiaLab desktop environment.

Batch Inference and Code Export

  • Batch Inference supports automated scoring across large datasets.
  • These workflows apply evidence from many records and compute posterior probabilities, classifications, or target distributions at scale.
  • BayesiaLab’s optional Code Export Module can generate static model code for R, SAS, PHP, VBA, Python, and JavaScript.
  • Exported code supports embedding trained Bayesian-network logic into external workflows and applications.

Bayesia Engine API

  • The Bayesia Engine API exposes key capabilities outside the graphical interface.
  • The Modeling Engine supports network construction and editing.
  • The Inference Engine supports automated and real-time inference workflows.
  • The Learning Engine provides programmatic access to discretization and learning algorithms.
  • APIs are delivered as Java class libraries (JAR files) for software integration.
  • Unlike static code export, the API allows applications to retain direct access to BayesiaLab’s modeling, learning, and inference capabilities.

Examples & Learn More