Convex-constrained Sparse Additive Modeling and Its Extensions

Junming Yin, Yaoliang Yu
Proceedings of the 33rd Conference on Uncertainty in Artificial Intelligence, PMLR R15:221-230, 2017.

Abstract

Sparse additive modeling is a class of effec- tive methods for performing high-dimensional nonparametric regression. In this work we show how shape constraints such as convex- ity/concavity and their extensions, can be in- tegrated into additive models. The proposed sparse difference of convex additive models (SDCAM) can estimate most continuous func- tions without any a priori smoothness assump- tion. Motivated by a characterization of dif- ference of convex functions, our method in- corporates a natural regularization functional to avoid overfitting and to reduce model com- plexity. Computationally, we develop an ef- ficient backfitting algorithm with linear per- iteration complexity. Experiments on both synthetic and real data confirm that our method is competitive against state-of-the-art sparse additive models, with improved performance in most scenarios.

Cite this Paper

BibTeX
@InProceedings{pmlr-vR15-yin17a, title = {Convex-constrained Sparse Additive Modeling and Its Extensions}, author = {Yin, Junming and Yu, Yaoliang}, booktitle = {Proceedings of the 33rd Conference on Uncertainty in Artificial Intelligence}, pages = {221--230}, year = {2017}, editor = {Elidan, Gal and Kersting, Kristian}, volume = {R15}, series = {Proceedings of Machine Learning Research}, month = {11--15 Aug}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/r15/main/assets/yin17a/yin17a.pdf}, url = {https://proceedings.mlr.press/r15/yin17a.html}, abstract = {Sparse additive modeling is a class of effec- tive methods for performing high-dimensional nonparametric regression. In this work we show how shape constraints such as convex- ity/concavity and their extensions, can be in- tegrated into additive models. The proposed sparse difference of convex additive models (SDCAM) can estimate most continuous func- tions without any a priori smoothness assump- tion. Motivated by a characterization of dif- ference of convex functions, our method in- corporates a natural regularization functional to avoid overfitting and to reduce model com- plexity. Computationally, we develop an ef- ficient backfitting algorithm with linear per- iteration complexity. Experiments on both synthetic and real data confirm that our method is competitive against state-of-the-art sparse additive models, with improved performance in most scenarios.}, note = {Reissued by PMLR on 04 October 2026.} }
Endnote
%0 Conference Paper %T Convex-constrained Sparse Additive Modeling and Its Extensions %A Junming Yin %A Yaoliang Yu %B Proceedings of the 33rd Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2017 %E Gal Elidan %E Kristian Kersting %F pmlr-vR15-yin17a %I PMLR %P 221--230 %U https://proceedings.mlr.press/r15/yin17a.html %V R15 %X Sparse additive modeling is a class of effec- tive methods for performing high-dimensional nonparametric regression. In this work we show how shape constraints such as convex- ity/concavity and their extensions, can be in- tegrated into additive models. The proposed sparse difference of convex additive models (SDCAM) can estimate most continuous func- tions without any a priori smoothness assump- tion. Motivated by a characterization of dif- ference of convex functions, our method in- corporates a natural regularization functional to avoid overfitting and to reduce model com- plexity. Computationally, we develop an ef- ficient backfitting algorithm with linear per- iteration complexity. Experiments on both synthetic and real data confirm that our method is competitive against state-of-the-art sparse additive models, with improved performance in most scenarios. %Z Reissued by PMLR on 04 October 2026.
APA
Yin, J. & Yu, Y.. (2017). Convex-constrained Sparse Additive Modeling and Its Extensions. Proceedings of the 33rd Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research R15:221-230 Available from https://proceedings.mlr.press/r15/yin17a.html. Reissued by PMLR on 04 October 2026.

Related Material