Scholarlink

The Kernel Feature Ordering by Conditional Independence (KFOCI) algorithm (Huang et al. J Mach Learn Res 23(216):1–58;2022. http://jmlr.org/papers/v23/21-493.html ) is a variable selection technique that does not assume a functional form for the model. In the present study, the KFOCI algorithm was implemented to search the variables that explain the spatial variation patterns of species diversity in a mountain system analyzed in Martínez-Camilo et al. (Biotropica 50(3):499–509;2018). We managed to obtain a subset of covariates that has smaller Bayesian Information Criterion (BIC) value when we use the parametric model implemented in Martínez-Camilo et al. (Biotropica 50(3):499–509;2018) than that obtained for the subset of covariates obtained in Martínez-Camilo et al. (Biotropica 50(3):499–509;2018). Both subsets have only climate variables, suggesting that climate variables are more important than the other factors in explaining diversity variation in mountain systems.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Kernel Feature Ordering by Conditional Independence in Species Diversity

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Gabriel A. Rodríguez-Yam,
  • José P. Sánchez-Venado

摘要

The Kernel Feature Ordering by Conditional Independence (KFOCI) algorithm (Huang et al. J Mach Learn Res 23(216):1–58;2022. http://jmlr.org/papers/v23/21-493.html ) is a variable selection technique that does not assume a functional form for the model. In the present study, the KFOCI algorithm was implemented to search the variables that explain the spatial variation patterns of species diversity in a mountain system analyzed in Martínez-Camilo et al. (Biotropica 50(3):499–509;2018). We managed to obtain a subset of covariates that has smaller Bayesian Information Criterion (BIC) value when we use the parametric model implemented in Martínez-Camilo et al. (Biotropica 50(3):499–509;2018) than that obtained for the subset of covariates obtained in Martínez-Camilo et al. (Biotropica 50(3):499–509;2018). Both subsets have only climate variables, suggesting that climate variables are more important than the other factors in explaining diversity variation in mountain systems.