Statistical Tests Based Multi-Strategy Learning Method to Infer the Gaussian Markov Network Structure
Inferring Gaussian Markov networks is an important and challenging problem in extensive fields, such as artificial intelligence, machine learning, and electronic commerce etc. However, most previous approaches can not extend to reconstruct large-scale networks with high efficiency, accuracy, as well as robustness. A Statistical Test based Structure Learning (STSL) algorithm for discovering Gaussian Markov network structure is proposed in this paper. The applicability and usefulness of the algorithm are demonstrated by both simulated and real data.