English

Sparse residual tree and forest

Numerical Analysis 2019-05-15 v1 Machine Learning Numerical Analysis

Abstract

Sparse residual tree (SRT) is an adaptive exploration method for multivariate scattered data approximation. It leads to sparse and stable approximations in areas where the data is sufficient or redundant, and points out the possible local regions where data refinement is needed. Sparse residual forest (SRF) is a combination of SRT predictors to further improve the approximation accuracy and stability according to the error characteristics of SRTs. The hierarchical parallel SRT algorithm is based on both tree decomposition and adaptive radial basis function (RBF) explorations, whereby for each child a sparse and proper RBF refinement is added to the approximation by minimizing the norm of the residual inherited from its parent. The convergence results are established for both SRTs and SRFs. The worst case time complexity of SRTs is O(Nlog2N)\mathcal{O}(N\log_2N) for the initial work and O(log2N)\mathcal{O}(\log_2N) for each prediction, meanwhile, the worst case storage requirement is O(Nlog2N)\mathcal{O}(N\log_2N), where the NN data points can be arbitrary distributed. Numerical experiments are performed for several illustrative examples.

Keywords

Cite

@article{arxiv.1902.06443,
  title  = {Sparse residual tree and forest},
  author = {Xin Xu and Xiaopeng Luo},
  journal= {arXiv preprint arXiv:1902.06443},
  year   = {2019}
}

Comments

21 pages, 9 figures

R2 v1 2026-06-23T07:43:26.311Z