A novel scalable, data efficient and correct Markov boundary learning algorithm under faithfulness condition
Sérgio Rodrigues de Morais, Alexandre Aussem · HAL (Le Centre pour la Communication Scientifique Directe) · 2008
In this paper, we discuss a novel scalable, data efficient and correct Markov boundary learning algorithm under faithfulness condition. The latter combines the main advantages of PCMB and IAMB yet avoids some of their drawbacks. An empiric evaluation of our algorithm is provided on synthetic and real sparse databases scaling up to 139,351 variables. Our method is shown to be efficient in terms of both runtime and accuracy.