A Minimalistic Approach to Sum-Product Network Learning for Real Applications
This work addresses the need for more robust SPN learning in real applications with messy data, though it is incremental as it builds on an existing algorithm.
The paper tackled the problem of making Sum-Product Network learning more practical by introducing MiniSPN, a simplified version of LearnSPN that handles missing data and heterogeneous features, achieving faster runtime and effective performance on benchmark and real-world datasets like Google's Knowledge Graph.
Sum-Product Networks (SPNs) are a class of expressive yet tractable hierarchical graphical models. LearnSPN is a structure learning algorithm for SPNs that uses hierarchical co-clustering to simultaneously identifying similar entities and similar features. The original LearnSPN algorithm assumes that all the variables are discrete and there is no missing data. We introduce a practical, simplified version of LearnSPN, MiniSPN, that runs faster and can handle missing data and heterogeneous features common in real applications. We demonstrate the performance of MiniSPN on standard benchmark datasets and on two datasets from Google's Knowledge Graph exhibiting high missingness rates and a mix of discrete and continuous features.