Scholarlink

<p>Exact interpretable learning is attractive in regulated decision settings, but solver runtime can vary substantially across datasets and solver families. We introduce structural meta-features derived from Feature Interaction Graphs (<span>fig</span> s) as interpretable signals for solver selection. We construct <span>fig</span> s from binarized tabular data using pairwise mutual information and extract topology-aware signatures such as density and estimated treewidth. Using a transparent shallow decision-tree selector, we demonstrate that <span>fig</span> features establish an interpretable structural view of solver behavior, complementing basic, statistical, and landmarking meta-features. Experiments on OpenML classification tasks show that topology-aware profiling exposes meaningful structural variation across datasets, although benchmark saturation prevents clear end-to-end routing gains over strong simple baselines. Our results validate <span>fig</span> as a principled, interpretable diagnostic tool for algorithm selection in exact learning; its diagnostic relevance becomes apparent on harder instances where solver runtime separation is substantial.</p>

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

Feature interaction graphs for exact interpretable learning solver selection: an empirical diagnostic study

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月27日

查看全文
  • Zhigao Huang,
  • Miao Pan,
  • Yuzhuo Pan,
  • Quanfa Li

摘要

Exact interpretable learning is attractive in regulated decision settings, but solver runtime can vary substantially across datasets and solver families. We introduce structural meta-features derived from Feature Interaction Graphs (fig s) as interpretable signals for solver selection. We construct fig s from binarized tabular data using pairwise mutual information and extract topology-aware signatures such as density and estimated treewidth. Using a transparent shallow decision-tree selector, we demonstrate that fig features establish an interpretable structural view of solver behavior, complementing basic, statistical, and landmarking meta-features. Experiments on OpenML classification tasks show that topology-aware profiling exposes meaningful structural variation across datasets, although benchmark saturation prevents clear end-to-end routing gains over strong simple baselines. Our results validate fig as a principled, interpretable diagnostic tool for algorithm selection in exact learning; its diagnostic relevance becomes apparent on harder instances where solver runtime separation is substantial.