TY - JOUR
T1 - A hierarchy of sum-product networks using robustness
AU - Conaty, Diarmaid
AU - Martínez del Rincon, Jesús
AU - de Campos, Cassio P.
PY - 2019/10
Y1 - 2019/10
N2 - Sum-product networks are a popular family of probabilistic graphical models that have been shown to achieve state-of-the-art performance in several tasks. When learning sum-product networks from scarce data, the obtained model may be prone to robustness issues, and where small variations of parameters could lead to different conclusions. We discuss the characteristics of sum-product networks as classifiers and study the robustness of them with respect to their parameters. Using a robustness measure to identify (possibly) unreliable decisions, we build a hierarchical approach where the classification task in testing time is deferred to another model if the outcome is deemed unreliable. We apply this approach on benchmark classification tasks across 47 datasets. Experiments show that the robustness measure can be a meaningful manner to build dynamic ensemble of classifiers and that our Hierarchical Sum-Product Network guarantees an improvement in accuracy.
AB - Sum-product networks are a popular family of probabilistic graphical models that have been shown to achieve state-of-the-art performance in several tasks. When learning sum-product networks from scarce data, the obtained model may be prone to robustness issues, and where small variations of parameters could lead to different conclusions. We discuss the characteristics of sum-product networks as classifiers and study the robustness of them with respect to their parameters. Using a robustness measure to identify (possibly) unreliable decisions, we build a hierarchical approach where the classification task in testing time is deferred to another model if the outcome is deemed unreliable. We apply this approach on benchmark classification tasks across 47 datasets. Experiments show that the robustness measure can be a meaningful manner to build dynamic ensemble of classifiers and that our Hierarchical Sum-Product Network guarantees an improvement in accuracy.
KW - Classification
KW - Robustness
KW - Sensitivity analysis
KW - Sum-product networks
UR - http://www.scopus.com/inward/record.url?scp=85073706601&partnerID=8YFLogxK
U2 - 10.1016/j.ijar.2019.07.007
DO - 10.1016/j.ijar.2019.07.007
M3 - Article
VL - 113
SP - 245
EP - 255
JO - International Journal of Approximate Reasoning
JF - International Journal of Approximate Reasoning
SN - 0888-613X
ER -