Machine Learning

Compositional Nonparametric Prediction: Statistical Efficiency and Greedy Regression Algorithm

Tagged: 

This topic contains 0 replies, has 1 voice, and was last updated by  arXiv 1 month ago.


  • arXiv
    5 pts

    Compositional Nonparametric Prediction: Statistical Efficiency and Greedy Regression Algorithm

    In this paper, we propose a compositional nonparametric method in which a model is expressed as a labeled binary tree of $2k+1$ nodes, where each node is either a summation, a multiplication, or the application of one of the $q$ basis functions to one of the $p$ covariates. We show that in order to recover a labeled binary tree from a given dataset, the sufficient number of samples is $O(klog(pq)+log(k!))$, and the necessary number of samples is $Omega(klog (pq)-log(k!))$. We further propose a greedy algorithm for regression in order to validate our theoretical findings through synthetic experiments.

    Compositional Nonparametric Prediction: Statistical Efficiency and Greedy Regression Algorithm
    by Yixi Xu, Jean Honorio, Xiao Wang
    https://arxiv.org/pdf/1704.01896v3.pdf

You must be logged in to reply to this topic.