Structural learning about independence graphs from multiple databases

In this paper, we propose an approach for structural learning of independence graphs from multiple databases or prior knowledge of conditional independencies. In our approach, we first learn a local graph from each database separately, and then we combine these local graphs together to construct a global graph over all variables. This approach can also be used in structural learning to utilize the prior knowledge of conditional independencies.Computer Science, Artificial IntelligenceEICPCI-S(ISTP)