Author(s)

Garret Vo^1*, Debdeep Pati²

¹Department of Industrial and Manufacturing Engineering, Florida State University, Tallahassee, FL, USA.
²Department of Statistics, Texas A and M University, College Station, TX, USA.

This paper presents a novel variable selection method in additive nonparametric regression model. This work is motivated by the need to select the number of nonparametric components and number of variables within each nonparametric component. The proposed method uses a combination of hard and soft shrinkages to separately control the number of additive components and the variables within each component. An efficient algorithm is developed to select the importance of variables and estimate the interaction network. Excellent performance is obtained in simulated and real data examples.

Additive, Gaussian Process, Interaction, Lasso, Sparsity, Variable Selection

Vo, G. and Pati, D. (2017) Sparse Additive Gaussian Process with Soft Interactions. Open Journal of Statistics, 7, 567-588. doi: 10.4236/ojs.2017.74039.