Interpretability has drawn increasing attention in machine learning. Partially linear additive models provide an attractive middle ground between the simplicity of generalized linear model and the ...