Algebraic Bayesian networks: Checking backbone connectivity

The paper investigates the construction of a joint graph as global structure network based on its primary structure, one problems arising in machine learning bases knowledge patterns with uncertainty, presented form algebraic Bayesian networks. aim research is to propose methods for solving inverse problem. As results, algorithms checking belonging family graphs and minimal are proposed, estimates their computational complexity made. An improved version special case an improvement general average also proposed algorithm membership graphs. problem recognition has not been previously researched; issue being addressed first time currently drafted. theoretical significance lies possibilities applying results further researches graph-theoretic invariants structures