Bayes Expert in GCN

The Bayes Expert Bayesian network software will facilitate the contribution of models that describe theories and data from the scientific community to the GCN, as well as play an important role in applying the data absorption technique to capture human body systems. Bayesian nets will be one method of presenting non-adaptive data to agents in order to prime the system for self-reinforcement.

Our Bayes Expert, a Bayesian network model creator, can populate the GCN repositories with real-world data from randomized controlled trials, the gold standard in science. Scientists can submit their medical literature search results, clinical trial results, and theories of causation supported by those results using the BayesExpert GUI in the SingularityNET marketplace. The BayesExpert dependency rule uses the statistical language of medical literature, including sensitivity and specificity, relative risk, and confidence intervals. These rules express dependent relations in the same way that a Bayesian net would, as translated from these statistics. Furthermore, the scientist can use standard logic's "and" and "or" functions, as well as other functions like "avg" (average) or "if-then-else" to more fully describe the relationships between variables.