Bayesian
Advantages of PredictionProbe’s BBN (PBBN)
By integrating Bayesian Belief Network (BBN) Technology and Probabilistic Technology, PredictionProbe has developed an innovative and powerful capability we call PBBN.
The use of PBBN along with our SPISE Bayesian Module provides the following valuable benefits and abilities:
- User-friendly, customizable design
- Dynamically imports states at runtime
- Dynamically imports Conditional Probably Tables (CPT) at runtime
- Output and report analysis data from individual nodes
- Automatically transfers analysis data from node to node
- Automatically produces outputs & reports network results
- Automatic CPT Generator:
- Ability to treat nodes/variables as continuous versus discrete
- Hidden nodes
- PDF/CDF and sensitivity data easily available for each node
- Automatically generate states
- Accurate CPT calculation with minimal number of states
- Parallel processing improves CPT calculation time
Probabilistic BBN is Used to Perform Bayesian Analysis in SPISE
